diff --git a/README.md b/README.md
index a05976d1f..3ab24fd32 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
 # AgentV
 
-Test AI targets on real repo tasks and measure what actually works.
+Test AI providers on real repo tasks and measure what actually works.
 
 ## Why?
 
@@ -10,18 +10,18 @@ Test AI targets on real repo tasks and measure what actually works.
 - **Version-controlled** — evals, judges, and results all live in Git
 - **Hybrid graders** — deterministic code checks + LLM-based subjective scoring
 - **CI/CD native** — exit codes, JSONL output, threshold flags for pipeline gating
-- **Any target** — run against agents, model providers, gateways, replay targets, CLI wrappers, transcript providers, and future app or service wrappers
+- **Any provider** — run against agents, model providers, gateways, replay providers, CLI wrappers, transcript providers, and future app or service wrappers
 
 ## Core Concepts
 
 - **Eval suite / tests** are the task corpus: the prompts, cases, datasets, and reusable field-local files you want to evaluate.
 - **Category** is derived from where the eval lives, such as folder path and file name. Use paths to organize the corpus instead of repeating category labels in every eval.
 - **Environment / fixtures / graders** are task-owned context: host or Docker setup, repos, setup scripts, files, fixtures, deterministic checks, and LLM grading prompts.
-- **Target** is the system under test: an agent, provider, gateway, replay target, CLI wrapper, transcript provider, or future app/service wrapper. Each eval selects one `target` by configured target `id` or with an eval-local target object.
-- **Tags** are run/result grouping labels. `tags.experiment` is the default experiment namespace, such as `with-skills` or `without-skills`; keep suite/category and target/model names out of that tag.
+- **Provider** is the configured system under test: an agent, model provider, gateway, replay provider, CLI wrapper, transcript provider, or future app/service wrapper. Each provider entry uses `id` for the backend/spec and optional `label` for the stable AgentV selection and result identity.
+- **Tags** are run/result grouping labels. `tags.experiment` is the default experiment namespace, such as `with-skills` or `without-skills`; keep suite/category and provider/model names out of that tag.
 - **Evaluate options** configure eval run behavior such as `max_concurrency`, repeat policy, and budgets.
 - **Default test** configures inherited per-test defaults such as score `threshold`.
-- **Run** is one concrete execution of a tagged eval against a resolved target that writes portable artifacts for readers such as Dashboard, compare, and trend.
+- **Run** is one concrete execution of a tagged eval against a resolved provider that writes portable artifacts for readers such as Dashboard, compare, and trend.
 
 ## Quick start
 
@@ -31,12 +31,12 @@ npm install -g agentv
 agentv init
 ```
 
-**2. Configure targets and graders** in `.agentv/config.yaml` — point to the system under test and the reusable grader. Provider settings live under `config`, and target `id` is the selection name used by evals and CLI flags:
+**2. Configure providers and graders** in `.agentv/providers.yaml` — point to the system under test and the reusable grader. Provider `id` names the backend/spec; `label` is the stable selection name used by evals and CLI flags:
 
 ```yaml
-targets:
-  - id: local-openai
-    provider: openai
+providers:
+  - id: openai
+    label: local-openai
     runtime: host
     config:
       api_format: chat
@@ -44,8 +44,8 @@ targets:
       api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
       model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
 
-  - id: local-openai-grader
-    provider: openai
+  - id: openai
+    label: local-openai-grader
     runtime: host
     config:
       api_format: chat
@@ -54,7 +54,7 @@ targets:
       model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
 
 defaults:
-  target: local-openai
+  provider: local-openai
   grader: local-openai-grader
 ```
 
@@ -82,7 +82,8 @@ options:
 description: Code generation quality
 tags:
   experiment: with-skills
-target: local-openai
+providers:
+  - local-openai
 evaluate_options:
   max_concurrency: 2
 
@@ -112,25 +113,25 @@ tests:
 Plain assertion strings are short-form rubric criteria: AgentV groups them into
 `llm-rubric` and writes grader detail to `grading.json.component_results` for
 the Dashboard. Use explicit `type: llm-rubric` when you need weights, required
-flags, `score_ranges`, a custom grader prompt, a grader target, or output
+flags, `score_ranges`, a custom grader prompt, a grader provider, or output
 transforms; use string `value` for free-form rubric checks. Executable graders
 use `type: script`.
 
-The target can be an eval-local object when this eval needs target settings of its own:
+The provider can be an eval-local object when this eval needs provider settings of its own:
 
 ```yaml
-description: Code generation quality with eval-local target settings
+description: Code generation quality with eval-local provider settings
 tags:
   experiment: with-skills
-target:
-  id: local-mini
-  provider: openai
-  runtime: host
-  config:
-    api_format: chat
-    base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
-    api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
-    model: gpt-5.4-mini
+providers:
+  - id: openai
+    label: local-mini
+    runtime: host
+    config:
+      api_format: chat
+      base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
+      api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
+      model: gpt-5.4-mini
 evaluate_options:
   repeat:
     count: 2
@@ -148,7 +149,7 @@ tests:
       input: Write FizzBuzz in Python
 ```
 
-`target: local-openai` resolves the configured target id from `.agentv/config.yaml` and uses its provider, model, hooks, and provider settings. The object form above defines a full eval-local target and must include enough provider configuration to run. AgentV records the resolved target information in run artifacts so results can be audited and replayed. The `tags.experiment` label stays `with-skills` because the condition is unchanged; the model/provider variation belongs to the resolved target metadata.
+`providers: [local-openai]` resolves the configured provider label from `.agentv/providers.yaml` and uses its backend, model, hooks, and provider settings. The object form above defines a full eval-local provider and must include enough provider configuration to run. AgentV records the resolved provider information in run artifacts so results can be audited and replayed. The `tags.experiment` label stays `with-skills` because the condition is unchanged; the model/provider variation belongs to the resolved provider metadata.
 
 Use `default_test.threshold` for the inherited per-test pass cutoff. `default_test` can also point at a shared file:
 
@@ -179,7 +180,7 @@ agentv results compare .agentv/results/<baseline-run-id>/.internal/index.jsonl .
 
 ## Results
 
-Each run writes a portable bundle directly under `.agentv/results/<run_id>/`. In this example, `tags.experiment: with-skills` names the condition being measured and `target: local-openai` selects the system under test from `.agentv/config.yaml`; both are recorded as metadata, not path segments. The `.internal/index.jsonl` file is the portable row index used by scripts, CI, and `agentv results compare`; per-case sidecars include the resolved eval and target configuration used for the run.
+Each run writes a portable bundle directly under `.agentv/results/<run_id>/`. In this example, `tags.experiment: with-skills` names the condition being measured and `providers: [local-openai]` selects the system under test from `.agentv/providers.yaml`; both are recorded as metadata, not path segments. The `.internal/index.jsonl` file is the portable row index used by scripts, CI, and `agentv results compare`; per-case sidecars include the resolved eval and provider configuration used for the run.
 
 ```bash
 agentv eval evals/my-eval.eval.yaml
@@ -192,7 +193,7 @@ Run bundle layout:
 .agentv/results/
 ├── 2026-06-30T08-30-00-000Z/     # <run_id> — one committed run bundle
 │   ├── summary.json              # run rollup: metadata, pass rate, counts, cost
-│   ├── fizzbuzz--a1b2c3d4/       # <result_dir> for one test/target row
+│   ├── fizzbuzz--a1b2c3d4/       # <result_dir> for one test/provider row
 │   │   ├── summary.json          # optional per-case rollup across samples
 │   │   ├── test/                 # generated test bundle: frozen inputs for reproducibility
 │   │   │   ├── EVAL.yaml         #   resolved eval spec
diff --git a/apps/cli/src/commands/create/commands.ts b/apps/cli/src/commands/create/commands.ts
index 02cf21488..8924e200d 100644
--- a/apps/cli/src/commands/create/commands.ts
+++ b/apps/cli/src/commands/create/commands.ts
@@ -42,7 +42,8 @@ export default defineAssertion(({ output }) => {
 
 const EVAL_TEMPLATES: Record<string, (name: string) => string> = {
   default: (name: string) => `description: ${name} evaluation suite
-target: default
+providers:
+  - default
 
 tests:
   - id: sample-test
@@ -54,7 +55,8 @@ tests:
         value: "well"
 `,
   rubric: (name: string) => `description: ${name} evaluation suite
-target: default
+providers:
+  - default
 
 tests:
   - id: sample-test
@@ -78,11 +80,11 @@ const PROVIDER_TEMPLATE = `#!/usr/bin/env bun
 /**
  * Custom provider scaffold.
  *
- * AgentV providers are configured via .agentv/targets.yaml using the CLI provider:
+ * AgentV providers are configured via .agentv/providers.yaml using the CLI provider:
  *
- *   targets:
- *     - name: my-target
- *       provider: cli
+ *   providers:
+ *     - id: cli
+ *       label: my-provider
  *       command: "bun run .agentv/providers/<name>.ts {PROMPT}"
  *
  * This script receives the prompt as a CLI argument and prints the response to stdout.
@@ -168,7 +170,7 @@ export const createProviderCommand = command({
     await writeFile(filePath, PROVIDER_TEMPLATE);
     console.log(`Created ${path.relative(process.cwd(), filePath)} (template: ${templateName})`);
     console.log(
-      `\nConfigure in .agentv/targets.yaml:\n  targets:\n    - name: ${name}\n      provider: cli\n      command: "bun run .agentv/providers/${name}.ts {PROMPT}"`,
+      `\nConfigure in .agentv/providers.yaml:\n  providers:\n    - id: cli\n      label: ${name}\n      command: "bun run .agentv/providers/${name}.ts {PROMPT}"`,
     );
   },
 });
diff --git a/apps/cli/src/commands/eval/commands/bundle.ts b/apps/cli/src/commands/eval/commands/bundle.ts
index 39816d153..44c43399c 100644
--- a/apps/cli/src/commands/eval/commands/bundle.ts
+++ b/apps/cli/src/commands/eval/commands/bundle.ts
@@ -67,9 +67,9 @@ function ensureTargetGraph(
         .map((entry) => entry.name)
         .sort()
         .join(', ');
-      const owner = requestedBy ? ` referenced by target '${requestedBy}'` : '';
+      const owner = requestedBy ? ` referenced by provider '${requestedBy}'` : '';
       throw new Error(
-        `Target '${name}'${owner} not found in ${targetsFilePath}. Available targets: ${available}`,
+        `Provider '${name}'${owner} not found in ${targetsFilePath}. Available providers: ${available}`,
       );
     }
     seen.add(name);
@@ -117,7 +117,7 @@ function definitionsWithEvalTargetSpec(
   if (!base) {
     const available = definitions.map((definition) => definition.name).join(', ');
     throw new Error(
-      `Target '${targetSpec.extends}' not found for eval-local target '${targetSpec.name}'. Available targets: ${available}`,
+      `Provider '${targetSpec.extends}' not found for eval-local provider '${targetSpec.name}'. Available providers: ${available}`,
     );
   }
   const effective = {
@@ -161,15 +161,37 @@ export const evalBundleCommand = command({
     target: multioption({
       type: array(string),
       long: 'target',
-      description: 'Target name to bundle (repeatable). Defaults to eval target(s) or default.',
+      description: '[Removed: use --provider <label>] Former provider selector',
     }),
     targets: option({
       type: optional(string),
       long: 'targets',
+      description: '[Removed: use --providers <path>] Former providers.yaml path',
+    }),
+    provider: multioption({
+      type: array(string),
+      long: 'provider',
+      description:
+        'Provider label to bundle (repeatable). Defaults to eval provider(s) or default.',
+    }),
+    providers: option({
+      type: optional(string),
+      long: 'providers',
       description: 'Path to providers.yaml (overrides discovery)',
     }),
   },
   handler: async (args) => {
+    if (args.target.length > 0) {
+      throw new Error(
+        `--target was removed from agentv eval bundle. Use --provider ${args.target[0]} instead.`,
+      );
+    }
+    if (args.targets !== undefined) {
+      throw new Error(
+        `--targets was removed from agentv eval bundle. Use --providers ${args.targets} instead.`,
+      );
+    }
+
     const cwd = process.cwd();
     const repoRoot = await findRepoRoot(cwd);
     const resolvedPaths = await resolveEvalPaths([args.evalPath], cwd);
@@ -192,10 +214,10 @@ export const evalBundleCommand = command({
     let targetNames: readonly string[];
     if (suite.inlineTarget) {
       definitions = [suite.inlineTarget];
-      targetNames = unique(args.target.length > 0 ? args.target : [suite.inlineTarget.name]);
+      targetNames = unique(args.provider.length > 0 ? args.provider : [suite.inlineTarget.name]);
     } else {
       const targetsFilePath = await discoverTargetsFile({
-        explicitPath: args.targets,
+        explicitPath: args.providers,
         testFilePath: evalFilePath,
         repoRoot,
         cwd,
@@ -209,8 +231,8 @@ export const evalBundleCommand = command({
       );
       const suiteTarget = await readTestSuiteTarget(evalFilePath);
       targetNames = unique(
-        args.target.length > 0
-          ? args.target
+        args.provider.length > 0
+          ? args.provider
           : (suite.targets ?? [suite.targetSpec?.name ?? suiteTarget ?? 'default']),
       );
       for (const targetName of targetNames) {
diff --git a/apps/cli/src/commands/eval/interactive.ts b/apps/cli/src/commands/eval/interactive.ts
index 63a9b48d4..4a442c96a 100644
--- a/apps/cli/src/commands/eval/interactive.ts
+++ b/apps/cli/src/commands/eval/interactive.ts
@@ -3,7 +3,7 @@ import path from 'node:path';
 import { listTargetNames, readTargetDefinitions } from '@agentv/core';
 import { checkbox, confirm, number, search, select } from '@inquirer/prompts';
 
-import { TARGET_FILE_CANDIDATES, fileExists } from '../../utils/targets.js';
+import { PROVIDER_FILE_CANDIDATES, fileExists } from '../../utils/targets.js';
 import {
   type DiscoveredEvalFile,
   discoverEvalFiles,
@@ -98,7 +98,7 @@ async function promptMainMenu(
       choices.push({
         name: '⏯  Resume last run',
         value: 'resume',
-        description: `${dirLabel} (target: ${lastConfig.target})`,
+        description: `${dirLabel} (provider: ${lastConfig.target})`,
       });
     }
   }
@@ -108,7 +108,7 @@ async function promptMainMenu(
     choices.push({
       name: '🔄 Rerun last config',
       value: 'rerun',
-      description: `${evalCount} eval file(s), target: ${lastConfig.target}`,
+      description: `${evalCount} eval file(s), provider: ${lastConfig.target}`,
     });
   }
 
@@ -143,7 +143,7 @@ async function promptNewEvaluation(cwd: string): Promise<InteractiveConfig | und
     return undefined;
   }
 
-  // Step 3: Select target
+  // Step 3: Select provider
   const target = await promptTargetSelection(cwd, selectedFiles[0].path);
 
   // Step 4: Advanced options
@@ -217,12 +217,12 @@ async function promptTargetSelection(cwd: string, firstEvalPath: string): Promis
   }
 
   if (targetNames.length === 1) {
-    console.log(`${ANSI_DIM}Using target: ${targetNames[0]}${ANSI_RESET}`);
+    console.log(`${ANSI_DIM}Using provider: ${targetNames[0]}${ANSI_RESET}`);
     return targetNames[0];
   }
 
   return search<string>({
-    message: 'Select a target (type to search)',
+    message: 'Select a provider (type to search)',
     source: async (term) => {
       const filtered = term
         ? targetNames.filter((t) => t.toLowerCase().includes(term.toLowerCase()))
@@ -232,7 +232,7 @@ async function promptTargetSelection(cwd: string, firstEvalPath: string): Promis
         return {
           name: t,
           value: t,
-          description: def ? `provider: ${def.provider}` : undefined,
+          description: def ? `backend: ${def.provider}` : undefined,
         };
       });
     },
@@ -263,7 +263,7 @@ async function findTargetsFile(
   }
 
   for (const dir of dirsToSearch) {
-    for (const candidate of TARGET_FILE_CANDIDATES) {
+    for (const candidate of PROVIDER_FILE_CANDIDATES) {
       const fullPath = `${dir}/${candidate}`;
       if (await fileExists(fullPath)) {
         return fullPath;
@@ -314,7 +314,7 @@ async function promptReviewAndConfirm(config: InteractiveConfig, cwd: string): P
   console.log(`\n${ANSI_BOLD}Review Configuration${ANSI_RESET}`);
   console.log(`${ANSI_DIM}${'─'.repeat(40)}${ANSI_RESET}`);
   console.log(`${ANSI_GREEN}Eval files:${ANSI_RESET}\n${evalDisplay}`);
-  console.log(`${ANSI_GREEN}Target:${ANSI_RESET}    ${config.target}`);
+  console.log(`${ANSI_GREEN}Provider:${ANSI_RESET}  ${config.target}`);
   console.log(`${ANSI_GREEN}Workers:${ANSI_RESET}   ${config.workers}`);
   console.log(`${ANSI_GREEN}Cache:${ANSI_RESET}     ${config.cache ? 'yes' : 'no'}`);
   console.log(`${ANSI_DIM}${'─'.repeat(40)}${ANSI_RESET}`);
diff --git a/apps/cli/src/commands/eval/run-eval.ts b/apps/cli/src/commands/eval/run-eval.ts
index 0a8e13e24..b1a7e5dc2 100644
--- a/apps/cli/src/commands/eval/run-eval.ts
+++ b/apps/cli/src/commands/eval/run-eval.ts
@@ -272,6 +272,8 @@ interface NormalizedOptions {
   readonly target?: string;
   readonly cliTargets: readonly string[];
   readonly targetsPath?: string;
+  /** Internal rerun-only carveout for generated test bundle targets.yaml artifacts. */
+  readonly allowLegacyTargetFiles?: boolean;
   readonly filter?: string | readonly string[];
   readonly workers?: number;
   /** --output <dir>: canonical artifact directory */
@@ -641,7 +643,7 @@ function normalizeOptions(
 
   const cliOutputDir = normalizeString(rawOptions.output);
 
-  // Normalize --target: can be a string (legacy) or string[] (multioption)
+  // Normalize provider selection: public --provider lowers into the historical internal target key.
   const rawTarget = rawOptions.target;
   let cliTargets: string[] = [];
   let singleTarget: string | undefined;
@@ -699,6 +701,7 @@ function normalizeOptions(
     target: singleTarget,
     cliTargets,
     targetsPath: normalizeString(rawOptions.targets),
+    allowLegacyTargetFiles: normalizeBoolean(rawOptions.allowLegacyTargetFiles),
     filter: normalizeFilter(rawOptions.filter),
     workers: workers > 0 ? workers : undefined,
     outputDir: cliOutputDir ?? configOutputDir,
@@ -1437,7 +1440,7 @@ async function prepareFileMetadata(params: {
       },
     ];
   } else {
-    // Determine target names: CLI --target flags override YAML
+    // Determine provider labels: CLI --provider flags override YAML
     const cliTargets = effectiveOptions.cliTargets;
     const experimentTargets = effectiveOptions.experimentTargets ?? [];
     const suiteTargetSpec = suite.targetSpec;
@@ -1471,6 +1474,7 @@ async function prepareFileMetadata(params: {
         repoRoot,
         cwd,
         explicitTargetsPath: effectiveOptions.targetsPath,
+        allowLegacyTargetFiles: effectiveOptions.allowLegacyTargetFiles,
         env: process.env,
         targetNames,
         targetRefs,
@@ -1490,6 +1494,7 @@ async function prepareFileMetadata(params: {
         repoRoot,
         cwd,
         explicitTargetsPath: effectiveOptions.targetsPath,
+        allowLegacyTargetFiles: effectiveOptions.allowLegacyTargetFiles,
         cliTargetName:
           targetSource === 'cli'
             ? targetNames.length === 1
@@ -1630,8 +1635,8 @@ async function runSingleEvalFile(params: {
   const explicitVariant = targetVariantForSelection(resolvedTargetSelection);
   const providerLabel = resolvedTargetSelection.resolvedTarget.kind;
   const targetMessage = options.verbose
-    ? `Using target (${resolvedTargetSelection.targetSource}): ${resolvedTargetSelection.targetName} ${buildTargetLabelSuffix(providerLabel, resolvedTargetSelection.resolvedTarget)} via ${resolvedTargetSelection.targetsFilePath}`
-    : `Using target: ${inlineTargetLabel}`;
+    ? `Using provider (${resolvedTargetSelection.targetSource}): ${resolvedTargetSelection.targetName} ${buildTargetLabelSuffix(providerLabel, resolvedTargetSelection.resolvedTarget)} via ${resolvedTargetSelection.targetsFilePath}`
+    : `Using provider: ${inlineTargetLabel}`;
   if (!progressReporter.isInteractive || options.verbose) {
     console.log(`${targetMessage}`);
   }
@@ -2754,7 +2759,7 @@ export async function runEvalCommand(
     // Suggest resume commands when execution errors are detected
     if (summary.executionErrorCount > 0 && !options.retryErrors && !options.resume) {
       const evalFileArgs = activeTestFiles.map((f) => path.relative(cwd, f)).join(' ');
-      const targetFlag = options.target ? ` --target ${options.target}` : '';
+      const targetFlag = options.target ? ` --provider ${options.target}` : '';
       const relativeRunDir = path.relative(cwd, runDir);
       console.log(
         `\nTip: ${summary.executionErrorCount} execution error(s) detected. Re-run failed tests with:\n` +
diff --git a/apps/cli/src/commands/eval/targets.ts b/apps/cli/src/commands/eval/targets.ts
index 8ad08b952..f112c856e 100644
--- a/apps/cli/src/commands/eval/targets.ts
+++ b/apps/cli/src/commands/eval/targets.ts
@@ -41,7 +41,7 @@ function resolveUseTarget(
   if (!current) {
     const available = listTargetNames(definitions).join(', ');
     throw new Error(
-      `Target '${name}' not found in ${targetsFilePath}. Available targets: ${available}`,
+      `Provider '${name}' not found in ${targetsFilePath}. Available providers: ${available}`,
     );
   }
 
@@ -60,7 +60,7 @@ function resolveUseTarget(
     if (!next) {
       const available = listTargetNames(definitions).join(', ');
       throw new Error(
-        `Target '${name}' use_target '${resolved.trim()}' not found in ${targetsFilePath}. Available targets: ${available}`,
+        `Provider '${name}' use_target '${resolved.trim()}' not found in ${targetsFilePath}. Available providers: ${available}`,
       );
     }
     current = next;
@@ -97,6 +97,7 @@ export interface TargetSelectionOptions {
   readonly repoRoot: string;
   readonly cwd: string;
   readonly explicitTargetsPath?: string;
+  readonly allowLegacyTargetFiles?: boolean;
   readonly cliTargetName?: string;
   readonly cliTargetNames?: readonly string[];
   readonly fileTargetName?: string;
@@ -159,14 +160,23 @@ function definitionsWithEffectiveTarget(
 }
 
 export async function selectTarget(options: TargetSelectionOptions): Promise<TargetSelection> {
-  const { testFilePath, repoRoot, cwd, explicitTargetsPath, cliTargetName, modelOverride, env } =
-    options;
+  const {
+    testFilePath,
+    repoRoot,
+    cwd,
+    explicitTargetsPath,
+    allowLegacyTargetFiles,
+    cliTargetName,
+    modelOverride,
+    env,
+  } = options;
 
   const targetsFilePath = await discoverTargetsFile({
     explicitPath: explicitTargetsPath,
     testFilePath,
     repoRoot,
     cwd,
+    allowLegacyTargetFiles,
   });
 
   // Validate the targets file and show warnings
@@ -195,7 +205,7 @@ export async function selectTarget(options: TargetSelectionOptions): Promise<Tar
       const message = useColors ? `${ANSI_RED}${error.message}${ANSI_RESET}` : error.message;
       console.error(`${prefix}${location} ${message}`);
     }
-    throw new Error(`Targets file validation failed with ${errors.length} error(s)`);
+    throw new Error(`Providers file validation failed with ${errors.length} error(s)`);
   }
 
   const definitions = await readTargetDefinitions(targetsFilePath);
@@ -233,7 +243,7 @@ export async function selectTarget(options: TargetSelectionOptions): Promise<Tar
     };
   } catch (error) {
     const message = error instanceof Error ? error.message : String(error);
-    throw new Error(`Failed to resolve target '${targetChoice.name}': ${message}`);
+    throw new Error(`Failed to resolve provider '${targetChoice.name}': ${message}`);
   }
 }
 
@@ -253,6 +263,7 @@ export async function selectMultipleTargets(
     repoRoot,
     cwd,
     explicitTargetsPath,
+    allowLegacyTargetFiles,
     env,
     targetNames,
     targetRefs,
@@ -278,6 +289,7 @@ export async function selectMultipleTargets(
     testFilePath,
     repoRoot,
     cwd,
+    allowLegacyTargetFiles,
   });
 
   // Validate targets file once
@@ -303,7 +315,7 @@ export async function selectMultipleTargets(
       const message = useColors ? `${ANSI_RED}${error.message}${ANSI_RESET}` : error.message;
       console.error(`${prefix}${location} ${message}`);
     }
-    throw new Error(`Targets file validation failed with ${errors.length} error(s)`);
+    throw new Error(`Providers file validation failed with ${errors.length} error(s)`);
   }
 
   const fileDefinitions = await readTargetDefinitions(targetsFilePath);
@@ -345,7 +357,7 @@ export async function selectMultipleTargets(
       });
     } catch (error) {
       const message = error instanceof Error ? error.message : String(error);
-      throw new Error(`Failed to resolve target '${name}': ${message}`);
+      throw new Error(`Failed to resolve provider '${name}': ${message}`);
     }
   }
 
diff --git a/apps/cli/src/commands/init/index.ts b/apps/cli/src/commands/init/index.ts
index faa05717d..418c929db 100644
--- a/apps/cli/src/commands/init/index.ts
+++ b/apps/cli/src/commands/init/index.ts
@@ -116,7 +116,7 @@ export async function initCommand(options: InitCommandOptions = {}): Promise<voi
   }
   console.log('\nYou can now:');
   console.log('  1. Copy .env.example to .env and add your API credentials');
-  console.log('  2. Configure targets in .agentv/targets.yaml');
+  console.log('  2. Configure providers in .agentv/providers.yaml');
   console.log('  3. Use AI skills to create and run evals');
   printSkillFirstInstructions();
 }
diff --git a/apps/cli/src/commands/pipeline/input.ts b/apps/cli/src/commands/pipeline/input.ts
index a54f01a13..33c3bdce3 100644
--- a/apps/cli/src/commands/pipeline/input.ts
+++ b/apps/cli/src/commands/pipeline/input.ts
@@ -1,5 +1,5 @@
 /**
- * `agentv pipeline input` — Extract eval inputs, target invocation info, and grader
+ * `agentv pipeline input` — Extract eval inputs, provider invocation info, and grader
  * configurations for subagent-mode eval runs.
  *
  * Reads an eval YAML file and writes a structured export directory that agents
@@ -67,15 +67,36 @@ export const evalInputCommand = command({
     target: option({
       type: optional(string),
       long: 'target',
-      description: 'Override target name from targets.yaml (mirrors eval run --target)',
+      description: '[Removed: use --provider <label>] Former provider selector',
     }),
     targets: option({
       type: optional(string),
       long: 'targets',
-      description: 'Path to targets.yaml (overrides discovery)',
+      description: '[Removed: use --providers <path>] Former providers.yaml path',
+    }),
+    provider: option({
+      type: optional(string),
+      long: 'provider',
+      description: 'Override provider label from providers.yaml',
+    }),
+    providers: option({
+      type: optional(string),
+      long: 'providers',
+      description: 'Path to providers.yaml (overrides discovery)',
     }),
   },
-  handler: async ({ evalPath, out, experiment, target, targets }) => {
+  handler: async ({ evalPath, out, experiment, target, targets, provider, providers }) => {
+    if (target !== undefined) {
+      throw new Error(
+        `--target was removed from agentv pipeline input. Use --provider ${target} instead.`,
+      );
+    }
+    if (targets !== undefined) {
+      throw new Error(
+        `--targets was removed from agentv pipeline input. Use --providers ${targets} instead.`,
+      );
+    }
+
     const resolvedEvalPath = resolve(evalPath);
     const outDir = resolve(out ?? buildDefaultRunDir(process.cwd(), experiment));
     const repoRoot = await findRepoRoot(dirname(resolvedEvalPath));
@@ -90,9 +111,9 @@ export const evalInputCommand = command({
       process.exit(1);
     }
 
-    // Try to resolve target for CLI invocation info.
+    // Try to resolve provider for CLI invocation info.
     // Non-CLI providers default to agent mode (executor subagents) unless
-    // subagent_mode_allowed: false is set in targets.yaml.
+    // subagent_mode_allowed: false is set in providers.yaml.
     let targetInfo: { kind: 'cli'; command: string; cwd: string; timeoutMs: number } | null = null;
     let targetName = 'agent';
     let targetKind = 'agent';
@@ -103,8 +124,8 @@ export const evalInputCommand = command({
         testFilePath: resolvedEvalPath,
         repoRoot,
         cwd: evalDir,
-        cliTargetName: target,
-        explicitTargetsPath: targets,
+        cliTargetName: provider,
+        explicitTargetsPath: providers,
         env: process.env,
       });
 
diff --git a/apps/cli/src/commands/pipeline/run.ts b/apps/cli/src/commands/pipeline/run.ts
index 2ad29bfaf..531e12108 100644
--- a/apps/cli/src/commands/pipeline/run.ts
+++ b/apps/cli/src/commands/pipeline/run.ts
@@ -1,13 +1,13 @@
 /**
- * `agentv pipeline run` — Combined command that runs input extraction, CLI target
+ * `agentv pipeline run` — Combined command that runs input extraction, CLI provider
  * invocation, and code grading in a single step.
  *
  * Equivalent to running:
  *   1. `agentv pipeline input <eval> --out <dir>`
- *   2. Invoking each CLI target in parallel (writing response.md + metrics.json)
+ *   2. Invoking each CLI provider in parallel (writing response.md + metrics.json)
  *   3. `agentv pipeline grade <dir>`
  *
- * For `kind: agent` targets, step 2 is skipped (subagent handles execution).
+ * For `kind: agent` providers, step 2 is skipped (subagent handles execution).
  *
  * To add new features: extend the handler — all logic is self-contained.
  */
@@ -64,7 +64,7 @@ function loadEnvFile(dir: string): Record<string, string> {
 export const evalRunCommand = command({
   name: 'run',
   description:
-    'Extract inputs, invoke CLI targets, and run script graders (for agent targets, use pipeline input + subagents)',
+    'Extract inputs, invoke CLI providers, and run script graders (for agent providers, use pipeline input + subagents)',
   args: {
     evalPath: positional({
       type: string,
@@ -79,7 +79,8 @@ export const evalRunCommand = command({
     workers: option({
       type: optional(number),
       long: 'workers',
-      description: 'Parallel workers for target invocation (default: targets.yaml workers, then 5)',
+      description:
+        'Parallel workers for provider invocation (default: providers.yaml workers, then 5)',
     }),
     experiment: option({
       type: optional(string),
@@ -95,15 +96,46 @@ export const evalRunCommand = command({
     target: option({
       type: optional(string),
       long: 'target',
-      description: 'Override target name from targets.yaml (mirrors eval run --target)',
+      description: '[Removed: use --provider <label>] Former provider selector',
     }),
     targets: option({
       type: optional(string),
       long: 'targets',
-      description: 'Path to targets.yaml (overrides discovery)',
+      description: '[Removed: use --providers <path>] Former providers.yaml path',
+    }),
+    provider: option({
+      type: optional(string),
+      long: 'provider',
+      description: 'Override provider label from providers.yaml',
+    }),
+    providers: option({
+      type: optional(string),
+      long: 'providers',
+      description: 'Path to providers.yaml (overrides discovery)',
     }),
   },
-  handler: async ({ evalPath, out, workers, experiment, graderType, target, targets }) => {
+  handler: async ({
+    evalPath,
+    out,
+    workers,
+    experiment,
+    graderType,
+    target,
+    targets,
+    provider,
+    providers,
+  }) => {
+    if (target !== undefined) {
+      throw new Error(
+        `--target was removed from agentv pipeline run. Use --provider ${target} instead.`,
+      );
+    }
+    if (targets !== undefined) {
+      throw new Error(
+        `--targets was removed from agentv pipeline run. Use --providers ${targets} instead.`,
+      );
+    }
+
     const resolvedEvalPath = resolve(evalPath);
     const outDir = resolve(out ?? buildDefaultRunDir(process.cwd(), experiment));
     const repoRoot = await findRepoRoot(dirname(resolvedEvalPath));
@@ -134,8 +166,8 @@ export const evalRunCommand = command({
         testFilePath: resolvedEvalPath,
         repoRoot,
         cwd: evalDir,
-        cliTargetName: target,
-        explicitTargetsPath: targets,
+        cliTargetName: provider,
+        explicitTargetsPath: providers,
         env: process.env,
       });
       targetName = selection.targetName;
diff --git a/apps/cli/src/commands/prepare/index.ts b/apps/cli/src/commands/prepare/index.ts
index 8ed3a57a4..6cef80519 100644
--- a/apps/cli/src/commands/prepare/index.ts
+++ b/apps/cli/src/commands/prepare/index.ts
@@ -401,9 +401,14 @@ export const prepareCommand = command({
       description: 'Exact test ID to prepare',
     }),
     target: option({
-      type: string,
+      type: optional(string),
       long: 'target',
-      description: 'Target name this prepared attempt is for',
+      description: '[Removed: use --provider <label>] Former provider selector',
+    }),
+    provider: option({
+      type: optional(string),
+      long: 'provider',
+      description: 'Provider label this prepared attempt is for',
     }),
     out: option({
       type: string,
@@ -416,8 +421,17 @@ export const prepareCommand = command({
       description: 'Output format: text (default) or json',
     }),
   },
-  handler: async ({ evalPath, testId, target, out, format }) => {
-    const result = await prepareAttempt({ evalPath, testId, target, outDir: out });
+  handler: async ({ evalPath, testId, target, provider, out, format }) => {
+    if (target !== undefined) {
+      throw new Error(
+        `--target was removed from agentv prepare. Use --provider ${target} instead.`,
+      );
+    }
+    if (!provider) {
+      throw new Error('--provider is required.');
+    }
+
+    const result = await prepareAttempt({ evalPath, testId, target: provider, outDir: out });
     if (format === 'json') {
       console.log(JSON.stringify(toCommandOutputWire(result), null, 2));
       return;
diff --git a/apps/cli/src/commands/results/eval-runner.ts b/apps/cli/src/commands/results/eval-runner.ts
index 3fb397621..159c44f99 100644
--- a/apps/cli/src/commands/results/eval-runner.ts
+++ b/apps/cli/src/commands/results/eval-runner.ts
@@ -4,7 +4,7 @@
  *
  * Provides Hono route handlers for:
  *   - GET  /api/eval/discover  — discover eval files in the project
- *   - GET  /api/eval/targets   — list available target names
+ *   - GET  /api/eval/targets   — list available provider labels
  *   - POST /api/eval/run       — launch an eval run as a child process
  *   - GET  /api/eval/status/:id — poll running eval status
  *   - GET  /api/eval/runs      — list active and recent Dashboard-launched runs
@@ -27,7 +27,7 @@ import { listTargetNames, readTargetDefinitions } from '@agentv/core';
 import type { Context } from 'hono';
 import type { Hono } from 'hono';
 
-import { TARGET_FILE_CANDIDATES } from '../../utils/targets.js';
+import { PROVIDER_FILE_CANDIDATES } from '../../utils/targets.js';
 import { discoverEvalFiles } from '../eval/discover.js';
 import {
   RESULT_INDEX_FILENAME,
@@ -42,7 +42,7 @@ interface DashboardRun {
   id: string;
   status: 'starting' | 'running' | 'finished' | 'failed';
   command: string;
-  /** Target name passed via --target (if any). Stored so the run list can show it before the first result is written. */
+  /** Provider label passed via --provider (if any). Stored so the run list can show it before the first result is written. */
   target?: string;
   /** Absolute path to the run directory (e.g. .agentv/results/<run_id>). Used to correlate this in-memory run with the filesystem run when the JSONL has 0 records yet. */
   outputDir?: string;
@@ -104,14 +104,14 @@ export function getActiveRunStatus(indexJsonlPath: string): DashboardRun['status
   return undefined;
 }
 
-// ── Discover targets file from project root ──────────────────────────────
+// ── Discover providers file from project root ────────────────────────────
 
 async function discoverTargetsInProject(cwd: string): Promise<readonly string[]> {
   const repoRoot = (await findRepoRoot(cwd)) ?? cwd;
 
-  // Try to find a targets file using the standard discovery
+  // Try to find a providers file using the standard discovery
   let targetsFilePath: string | undefined;
-  for (const candidate of TARGET_FILE_CANDIDATES) {
+  for (const candidate of PROVIDER_FILE_CANDIDATES) {
     const fullPath = path.join(cwd, candidate);
     if (existsSync(fullPath)) {
       targetsFilePath = fullPath;
@@ -119,7 +119,7 @@ async function discoverTargetsInProject(cwd: string): Promise<readonly string[]>
     }
   }
   if (!targetsFilePath) {
-    for (const candidate of TARGET_FILE_CANDIDATES) {
+    for (const candidate of PROVIDER_FILE_CANDIDATES) {
       const fullPath = path.join(repoRoot, candidate);
       if (existsSync(fullPath)) {
         targetsFilePath = fullPath;
@@ -206,9 +206,9 @@ function buildCliArgs(req: RunEvalRequest, experiment?: string): string[] {
     }
   }
 
-  // Target override
+  // Provider override
   if (req.target?.trim()) {
-    args.push('--target', req.target.trim());
+    args.push('--provider', req.target.trim());
   }
 
   if (experiment && req.experiment?.trim()) {
diff --git a/apps/cli/src/commands/runs/rerun.ts b/apps/cli/src/commands/runs/rerun.ts
index 7471b7718..7b3fdcf0c 100644
--- a/apps/cli/src/commands/runs/rerun.ts
+++ b/apps/cli/src/commands/runs/rerun.ts
@@ -26,6 +26,7 @@ import { type ResultManifestRecord, parseResultManifest } from '../results/manif
 
 const TASK_EVAL_FILENAME = 'EVAL.yaml';
 const TASK_PROVIDERS_FILENAME = 'providers.yaml';
+const TASK_TARGETS_FILENAME = 'targets.yaml';
 const ENV_REF_PATTERN = /\$\{\{\s*([A-Za-z_][A-Za-z0-9_]*)\s*\}\}/g;
 
 interface SelectedTaskBundle {
@@ -332,11 +333,19 @@ async function loadSelectedTaskBundles(options: {
         options.sourceRunDir,
         bundleDir && `${bundleDir}/${TASK_EVAL_FILENAME}`,
       );
+    const rawRecord = record as unknown as Record<string, unknown>;
+    const legacyTargetsPath =
+      typeof rawRecord.targets_path === 'string' ? rawRecord.targets_path : undefined;
     const providersPath =
       resolveRelativeRunPath(options.sourceRunDir, record.providers_path) ??
+      resolveRelativeRunPath(options.sourceRunDir, legacyTargetsPath) ??
       resolveRelativeRunPath(
         options.sourceRunDir,
         bundleDir && `${bundleDir}/${TASK_PROVIDERS_FILENAME}`,
+      ) ??
+      resolveRelativeRunPath(
+        options.sourceRunDir,
+        bundleDir && `${bundleDir}/${TASK_TARGETS_FILENAME}`,
       );
     const testDir =
       resolveRelativeRunPath(options.sourceRunDir, bundleDir) ??
@@ -511,6 +520,8 @@ export const runsRerunCommand = command({
         workers: args.workers,
         dryRun: args.dryRun,
         verbose: args.verbose,
+        // Legacy generated run bundles stored the captured provider graph as targets.yaml.
+        allowLegacyTargetFiles: true,
         sourceMetadataByEvalFile: buildSourceMetadataByEvalFile(sourceRunDir, indexPath, selected),
       },
     });
diff --git a/apps/cli/src/templates/.agentv/providers.yaml b/apps/cli/src/templates/.agentv/providers.yaml
index c19c9082d..1f9effcc2 100644
--- a/apps/cli/src/templates/.agentv/providers.yaml
+++ b/apps/cli/src/templates/.agentv/providers.yaml
@@ -1,37 +1,27 @@
-# A list of all supported evaluation providers for the project.
-# In each provider, `id` names the backend/spec and `label` is the stable
-# AgentV selection identity used by --provider and eval `providers`.
-# Actual values for paths/keys are stored in the local .env file.
-
 providers:
   - id: azure
     label: default
     endpoint: "{{ env.AZURE_OPENAI_ENDPOINT }}"
     api_key: "{{ env.AZURE_OPENAI_API_KEY }}"
     model: "{{ env.AZURE_DEPLOYMENT_NAME }}"
-
   - id: codex
     label: codex
     cwd: "{{ env.CODEX_WORKSPACE_DIR }}"
     log_dir: "{{ env.CODEX_LOG_DIR }}"
     stream_log: raw
-
   - id: claude
     label: claude
     stream_log: raw
-
   - id: azure
     label: azure-llm
     endpoint: "{{ env.AZURE_OPENAI_ENDPOINT }}"
     api_key: "{{ env.AZURE_OPENAI_API_KEY }}"
     model: "{{ env.AZURE_DEPLOYMENT_NAME }}"
     version: "{{ env.AZURE_OPENAI_API_VERSION }}"
-
   - id: gemini
     label: gemini-llm
     api_key: "{{ env.GOOGLE_GENERATIVE_AI_API_KEY }}"
     model: "{{ env.GEMINI_MODEL_NAME }}"
-
   - id: cli
     label: local_cli
     command: uv run ./mock_cli.py --prompt {PROMPT} {FILES} --output {OUTPUT_FILE}
@@ -39,17 +29,14 @@ providers:
     cwd: "{{ env.CLI_EVALS_DIR }}"
     healthcheck:
       command: uv run ./mock_cli.py --healthcheck
-
   - id: openrouter
     label: mimo
     api_key: "{{ env.OPENROUTER_API_KEY }}"
     model: xiaomi/mimo-v2.5-pro
-
   - id: openrouter
     label: mimo-flash
     api_key: "{{ env.OPENROUTER_API_KEY }}"
     model: xiaomi/mimo-v2-flash
-
 defaults:
   provider: default
   grader: azure-llm
diff --git a/apps/cli/src/templates/.env.example b/apps/cli/src/templates/.env.example
index 1f8e22057..b6a3aa71a 100644
--- a/apps/cli/src/templates/.env.example
+++ b/apps/cli/src/templates/.env.example
@@ -18,5 +18,5 @@ GEMINI_MODEL_NAME=gemini-3-flash-preview
 # Anthropic
 ANTHROPIC_API_KEY=your-anthropic-api-key-here
 
-# CLI provider sample (used by the local_cli target)
-CLI_EVALS_DIR=./docs/examples/simple/evals/local-cli
\ No newline at end of file
+# CLI provider sample (used by the local_cli provider)
+CLI_EVALS_DIR=./docs/examples/simple/evals/local-cli
diff --git a/apps/cli/src/utils/targets.ts b/apps/cli/src/utils/targets.ts
index 45502f6b9..7324c2d72 100644
--- a/apps/cli/src/utils/targets.ts
+++ b/apps/cli/src/utils/targets.ts
@@ -1,15 +1,22 @@
 import { constants } from 'node:fs';
-import { access } from 'node:fs/promises';
+import { access, stat } from 'node:fs/promises';
 import path from 'node:path';
 import { buildDirectoryChain } from '@agentv/core';
 
-export const TARGET_FILE_CANDIDATES = [
+export const PROVIDER_FILE_CANDIDATES = [
   'providers.yaml',
   'providers.yml',
   path.join('.agentv', 'providers.yaml'),
   path.join('.agentv', 'providers.yml'),
 ] as const;
 
+const LEGACY_TARGET_FILE_CANDIDATES = [
+  'targets.yaml',
+  'targets.yml',
+  path.join('.agentv', 'targets.yaml'),
+  path.join('.agentv', 'targets.yml'),
+] as const;
+
 export async function fileExists(filePath: string): Promise<boolean> {
   try {
     await access(filePath, constants.F_OK);
@@ -19,24 +26,81 @@ export async function fileExists(filePath: string): Promise<boolean> {
   }
 }
 
+async function pathKind(filePath: string): Promise<'file' | 'directory' | undefined> {
+  try {
+    const info = await stat(filePath);
+    if (info.isDirectory()) {
+      return 'directory';
+    }
+    if (info.isFile()) {
+      return 'file';
+    }
+    return undefined;
+  } catch {
+    return undefined;
+  }
+}
+
+function isLegacyTargetFile(filePath: string): boolean {
+  return /^targets\.ya?ml$/i.test(path.basename(filePath));
+}
+
+function removedTargetsFileError(filePath: string): Error {
+  return new Error(
+    `Authored targets.yaml files were removed. Rename ${filePath} to providers.yaml and use --providers to specify it explicitly.`,
+  );
+}
+
+async function findProviderFileInDirectory(directory: string): Promise<string | undefined> {
+  for (const candidate of PROVIDER_FILE_CANDIDATES) {
+    const fullPath = path.join(directory, candidate);
+    if ((await pathKind(fullPath)) === 'file') {
+      return fullPath;
+    }
+  }
+  return undefined;
+}
+
+async function findLegacyTargetFileInDirectory(directory: string): Promise<string | undefined> {
+  for (const candidate of LEGACY_TARGET_FILE_CANDIDATES) {
+    const fullPath = path.join(directory, candidate);
+    if ((await pathKind(fullPath)) === 'file') {
+      return fullPath;
+    }
+  }
+  return undefined;
+}
+
 export async function discoverTargetsFile(options: {
   readonly explicitPath?: string;
   readonly testFilePath: string;
   readonly repoRoot: string;
   readonly cwd: string;
+  readonly allowLegacyTargetFiles?: boolean;
 }): Promise<string> {
-  const { explicitPath, testFilePath, repoRoot, cwd } = options;
+  const { explicitPath, testFilePath, repoRoot, cwd, allowLegacyTargetFiles = false } = options;
 
   if (explicitPath) {
     const resolvedExplicit = path.resolve(explicitPath);
-    if (await fileExists(resolvedExplicit)) {
+    const kind = await pathKind(resolvedExplicit);
+    if (kind === 'file') {
+      if (!allowLegacyTargetFiles && isLegacyTargetFile(resolvedExplicit)) {
+        throw removedTargetsFileError(resolvedExplicit);
+      }
       return resolvedExplicit;
     }
 
-    for (const candidate of TARGET_FILE_CANDIDATES) {
-      const nested = path.join(resolvedExplicit, candidate);
-      if (await fileExists(nested)) {
-        return nested;
+    if (kind === 'directory') {
+      const providerFile = await findProviderFileInDirectory(resolvedExplicit);
+      if (providerFile) {
+        return providerFile;
+      }
+      const legacyTargetFile = await findLegacyTargetFileInDirectory(resolvedExplicit);
+      if (legacyTargetFile) {
+        if (allowLegacyTargetFiles) {
+          return legacyTargetFile;
+        }
+        throw removedTargetsFileError(legacyTargetFile);
       }
     }
 
@@ -52,11 +116,19 @@ export async function discoverTargetsFile(options: {
   }
 
   for (const directory of directories) {
-    for (const candidate of TARGET_FILE_CANDIDATES) {
-      const fullPath = path.join(directory, candidate);
-      if (await fileExists(fullPath)) {
-        return fullPath;
+    const providerFile = await findProviderFileInDirectory(directory);
+    if (providerFile) {
+      return providerFile;
+    }
+  }
+
+  for (const directory of directories) {
+    const legacyTargetFile = await findLegacyTargetFileInDirectory(directory);
+    if (legacyTargetFile) {
+      if (allowLegacyTargetFiles) {
+        return legacyTargetFile;
       }
+      throw removedTargetsFileError(legacyTargetFile);
     }
   }
 
diff --git a/apps/cli/test/commands/eval/bundle.test.ts b/apps/cli/test/commands/eval/bundle.test.ts
index 4f45384ef..20fd12ff0 100644
--- a/apps/cli/test/commands/eval/bundle.test.ts
+++ b/apps/cli/test/commands/eval/bundle.test.ts
@@ -106,7 +106,7 @@ tests: ../data/cases.yaml
     return { sourceDir, bundleDir, evalPath, sourceEvalBefore };
   }
 
-  it('bundles inherited targets, relative data, environment workdirs, and scripts into a runnable directory', async () => {
+  it('bundles inherited targets, relative data, environment workdirs, and scripts into a portable directory', async () => {
     const { sourceDir, bundleDir, evalPath, sourceEvalBefore } =
       await createPortableSourceFixture();
 
@@ -260,4 +260,58 @@ tests:
     expect(output).toContain('environment.workdir for test "missing-template"');
     expect(output).toContain('not found');
   }, 30_000);
+
+  it('hard-errors removed target-named bundle flags with provider guidance', async () => {
+    const sourceDir = path.join(tempDir, 'legacy-flags-source');
+    const bundleDir = path.join(tempDir, 'legacy-flags-bundle');
+    await mkdir(path.join(sourceDir, '.agentv'), { recursive: true });
+    await mkdir(path.join(sourceDir, 'evals'), { recursive: true });
+    await writeFile(
+      path.join(sourceDir, '.agentv', 'providers.yaml'),
+      `providers:
+  - id: mock
+    label: default
+`,
+      'utf8',
+    );
+    await writeFile(
+      path.join(sourceDir, 'evals', 'suite.eval.yaml'),
+      `prompts:
+  - "{{ input }}"
+tests:
+  - id: case
+    vars:
+      input: hello
+`,
+      'utf8',
+    );
+
+    const targetResult = await runCli(sourceDir, [
+      'eval',
+      'bundle',
+      'evals/suite.eval.yaml',
+      '--out',
+      bundleDir,
+      '--target',
+      'default',
+    ]);
+    expect(targetResult.exitCode).toBe(1);
+    expect(`${targetResult.stdout}\n${targetResult.stderr}`).toContain(
+      '--target was removed from agentv eval bundle. Use --provider default instead.',
+    );
+
+    const targetsResult = await runCli(sourceDir, [
+      'eval',
+      'bundle',
+      'evals/suite.eval.yaml',
+      '--out',
+      bundleDir,
+      '--targets',
+      '.agentv/targets.yaml',
+    ]);
+    expect(targetsResult.exitCode).toBe(1);
+    expect(`${targetsResult.stdout}\n${targetsResult.stderr}`).toContain(
+      '--targets was removed from agentv eval bundle. Use --providers .agentv/targets.yaml instead.',
+    );
+  }, 30_000);
 });
diff --git a/apps/cli/test/commands/eval/targets.test.ts b/apps/cli/test/commands/eval/targets.test.ts
index 99bc50ad1..1fb5c21fc 100644
--- a/apps/cli/test/commands/eval/targets.test.ts
+++ b/apps/cli/test/commands/eval/targets.test.ts
@@ -64,4 +64,81 @@ describe('eval target selection', () => {
     expect(selections[0]?.targetLabel).toBeUndefined();
     expect(selections[0]?.resolvedTarget.kind).toBe('mock');
   });
+
+  it('prefers providers.yaml over legacy targets.yaml during provider discovery', async () => {
+    const agentvDir = path.join(tempDir, '.agentv');
+    await mkdir(agentvDir, { recursive: true });
+    await writeFile(
+      path.join(agentvDir, 'providers.yaml'),
+      ['providers:', '  - id: mock', '    label: modern', '    response: modern', ''].join('\n'),
+    );
+    await writeFile(
+      path.join(agentvDir, 'targets.yaml'),
+      ['providers:', '  - id: mock', '    label: legacy', '    response: legacy', ''].join('\n'),
+    );
+    const evalPath = path.join(tempDir, 'provider-discovery.eval.yaml');
+    await writeFile(
+      evalPath,
+      [
+        'providers:',
+        '  - modern',
+        'prompts:',
+        '  - "{{ input }}"',
+        'tests:',
+        '  - id: provider-case',
+        '    vars:',
+        '      input: hello',
+      ].join('\n'),
+    );
+
+    const suite = await loadTestSuite(evalPath, tempDir);
+    const selections = await selectMultipleTargets({
+      testFilePath: evalPath,
+      repoRoot: tempDir,
+      cwd: tempDir,
+      env: {},
+      targetNames: suite.targets ?? [],
+      targetRefs: suite.targetRefs,
+      targetSource: 'test-file',
+    });
+
+    expect(selections[0]?.targetName).toBe('modern');
+    expect(selections[0]?.resolvedTarget.config.response).toBe('modern');
+  });
+
+  it('hard-rejects legacy targets.yaml as authored provider config', async () => {
+    const agentvDir = path.join(tempDir, '.agentv');
+    await mkdir(agentvDir, { recursive: true });
+    await writeFile(
+      path.join(agentvDir, 'targets.yaml'),
+      ['providers:', '  - id: mock', '    label: legacy', ''].join('\n'),
+    );
+    const evalPath = path.join(tempDir, 'legacy-targets.eval.yaml');
+    await writeFile(
+      evalPath,
+      [
+        'providers:',
+        '  - legacy',
+        'prompts:',
+        '  - "{{ input }}"',
+        'tests:',
+        '  - id: provider-case',
+        '    vars:',
+        '      input: hello',
+      ].join('\n'),
+    );
+
+    const suite = await loadTestSuite(evalPath, tempDir);
+    await expect(
+      selectMultipleTargets({
+        testFilePath: evalPath,
+        repoRoot: tempDir,
+        cwd: tempDir,
+        env: {},
+        targetNames: suite.targets ?? [],
+        targetRefs: suite.targetRefs,
+        targetSource: 'test-file',
+      }),
+    ).rejects.toThrow(/Authored targets\.yaml files were removed.*providers\.yaml/);
+  });
 });
diff --git a/apps/cli/test/commands/grade/grade-prepared.test.ts b/apps/cli/test/commands/grade/grade-prepared.test.ts
index 97e466f75..9ff71fd6b 100644
--- a/apps/cli/test/commands/grade/grade-prepared.test.ts
+++ b/apps/cli/test/commands/grade/grade-prepared.test.ts
@@ -139,7 +139,7 @@ describe('agentv grade prepared attempts', () => {
     const runDir = path.join(tempDir, 'runs', 'prepared-grade');
 
     await runCli(
-      ['prepare', evalPath, '--test-id', 'case-1', '--target', 'codex', '--out', preparedDir],
+      ['prepare', evalPath, '--test-id', 'case-1', '--provider', 'codex', '--out', preparedDir],
       tempDir,
     );
     await writeFile(
@@ -272,7 +272,7 @@ describe('agentv grade prepared attempts', () => {
     const runDir = path.join(tempDir, 'runs', 'trace-tools');
 
     await runCli(
-      ['prepare', evalPath, '--test-id', 'case-1', '--target', 'codex', '--out', preparedDir],
+      ['prepare', evalPath, '--test-id', 'case-1', '--provider', 'codex', '--out', preparedDir],
       tempDir,
     );
     await writeFile(
@@ -331,7 +331,7 @@ describe('agentv grade prepared attempts', () => {
     const tracePath = path.join(tempDir, 'trace.jsonl');
 
     await runCli(
-      ['prepare', evalPath, '--test-id', 'case-1', '--target', 'codex', '--out', preparedDir],
+      ['prepare', evalPath, '--test-id', 'case-1', '--provider', 'codex', '--out', preparedDir],
       tempDir,
     );
     await writeFile(
@@ -447,7 +447,7 @@ describe('agentv grade prepared attempts', () => {
     const tracePath = path.join(tempDir, 'wrong-transcript-target.jsonl');
 
     await runCli(
-      ['prepare', evalPath, '--test-id', 'case-1', '--target', 'codex', '--out', preparedDir],
+      ['prepare', evalPath, '--test-id', 'case-1', '--provider', 'codex', '--out', preparedDir],
       tempDir,
     );
     await writeFile(
@@ -491,7 +491,7 @@ describe('agentv grade prepared attempts', () => {
     const tracePath = path.join(tempDir, 'wrong-envelope-target.json');
 
     await runCli(
-      ['prepare', evalPath, '--test-id', 'case-1', '--target', 'codex', '--out', preparedDir],
+      ['prepare', evalPath, '--test-id', 'case-1', '--provider', 'codex', '--out', preparedDir],
       tempDir,
     );
     await writeFile(
diff --git a/apps/cli/test/commands/prepare/prepare.test.ts b/apps/cli/test/commands/prepare/prepare.test.ts
index b7f719d4b..6d15d800b 100644
--- a/apps/cli/test/commands/prepare/prepare.test.ts
+++ b/apps/cli/test/commands/prepare/prepare.test.ts
@@ -131,7 +131,7 @@ describe('agentv prepare', () => {
         evalPath,
         '--test-id',
         'case-1',
-        '--target',
+        '--provider',
         'codex',
         '--out',
         outDir,
@@ -205,7 +205,7 @@ describe('agentv prepare', () => {
         evalPath,
         '--test-id',
         'case-1',
-        '--target',
+        '--provider',
         'codex',
         '--out',
         outDir,
@@ -288,7 +288,7 @@ tests:
         evalPath,
         '--test-id',
         'case-1',
-        '--target',
+        '--provider',
         'codex',
         '--out',
         outDir,
diff --git a/apps/cli/test/commands/runs/rerun.test.ts b/apps/cli/test/commands/runs/rerun.test.ts
index 1580312fd..1d42c9476 100644
--- a/apps/cli/test/commands/runs/rerun.test.ts
+++ b/apps/cli/test/commands/runs/rerun.test.ts
@@ -39,6 +39,7 @@ async function writeTaskBundle(options: {
   readonly testId: string;
   readonly targetsYaml: string;
   readonly legacyTaskDir?: boolean;
+  readonly legacyTargetsPath?: boolean;
 }): Promise<Record<string, unknown>> {
   const artifactDir = path.join(options.sourceRunDir, options.testId);
   const bundleDirname = options.legacyTaskDir ? 'task' : 'test';
@@ -66,7 +67,8 @@ tests:
 `,
     'utf8',
   );
-  await writeFile(path.join(bundleDir, 'providers.yaml'), options.targetsYaml, 'utf8');
+  const providersFilename = options.legacyTargetsPath ? 'targets.yaml' : 'providers.yaml';
+  await writeFile(path.join(bundleDir, providersFilename), options.targetsYaml, 'utf8');
   await writeFile(path.join(artifactDir, 'grading.json'), '{"assertions":[]}\n', 'utf8');
   await writeFile(path.join(artifactDir, 'timing.json'), '{"duration_ms":1}\n', 'utf8');
   await writeFile(path.join(outputsDir, 'answer.md'), '@[assistant]:\nCaptured answer\n', 'utf8');
@@ -74,7 +76,8 @@ tests:
   const bundlePaths = {
     [`${options.legacyTaskDir ? 'task' : 'test'}_dir`]: `${options.testId}/${bundleDirname}`,
     eval_path: `${options.testId}/${bundleDirname}/EVAL.yaml`,
-    providers_path: `${options.testId}/${bundleDirname}/providers.yaml`,
+    [options.legacyTargetsPath ? 'targets_path' : 'providers_path']:
+      `${options.testId}/${bundleDirname}/${providersFilename}`,
   };
 
   return {
@@ -93,7 +96,7 @@ tests:
 
 async function createBundleFixture(
   targetsYaml = DEFAULT_TARGETS,
-  options?: { readonly legacyTaskDir?: boolean },
+  options?: { readonly legacyTaskDir?: boolean; readonly legacyTargetsPath?: boolean },
 ): Promise<BundleFixture> {
   const baseDir = await mkdtemp(path.join(tmpdir(), 'agentv-rerun-'));
   const cwd = path.join(baseDir, 'workspace');
@@ -108,12 +111,14 @@ async function createBundleFixture(
       testId: 'case-alpha',
       targetsYaml,
       legacyTaskDir: options?.legacyTaskDir,
+      legacyTargetsPath: options?.legacyTargetsPath,
     }),
     await writeTaskBundle({
       sourceRunDir,
       testId: 'case-beta',
       targetsYaml,
       legacyTaskDir: options?.legacyTaskDir,
+      legacyTargetsPath: options?.legacyTargetsPath,
     }),
   ];
   await writeFile(
@@ -280,6 +285,24 @@ describe('agentv runs rerun', () => {
     });
   }, 30_000);
 
+  it('reruns legacy generated targets_path bundles for backward compatibility', async () => {
+    const created = await createBundleFixture(DEFAULT_TARGETS, { legacyTargetsPath: true });
+
+    const result = await runCli(created, [
+      'runs',
+      'rerun',
+      created.sourceRunDir,
+      '--test-id',
+      'case-alpha',
+      '--output',
+      created.outputDir,
+    ]);
+
+    expect(result.exitCode).toBe(0);
+    const { rows } = await readOutputBundle(created.outputDir);
+    expect(rows.map((row) => row.test_id)).toEqual(['case-alpha']);
+  }, 30_000);
+
   it('fails clearly for missing env and accepts an explicit env file', async () => {
     const created = await fixture(`providers:
   - id: cli
@@ -296,7 +319,7 @@ describe('agentv runs rerun', () => {
       '--dry-run',
     ]);
     expect(missing.exitCode).toBe(1);
-    expect(missing.stderr).toContain("Failed to resolve target 'captured'");
+    expect(missing.stderr).toContain("Failed to resolve provider 'captured'");
     expect(missing.stderr).toContain('LOCAL_AGENT_COMMAND');
 
     const withAmbientEnv = await runCli(
diff --git a/apps/cli/test/eval.integration.test.ts b/apps/cli/test/eval.integration.test.ts
index 6777a3495..f748167cc 100644
--- a/apps/cli/test/eval.integration.test.ts
+++ b/apps/cli/test/eval.integration.test.ts
@@ -287,7 +287,7 @@ describe('agentv eval CLI', () => {
       const { stdout } = await runCli(fixture, ['eval', fixture.testFilePath, '--verbose']);
 
       // Don't check stderr - it may contain stack traces or other diagnostics
-      expect(stdout).toContain('Using target (test-file): file-target [provider=mock]');
+      expect(stdout).toContain('Using provider (test-file): file-target [provider=mock]');
       expect(stdout).toContain('Mean score: 75%');
       // Std deviation is an implementation detail - don't check it
 
diff --git a/apps/web/src/content/docs/docs/v4.42.4/evaluation/batch-cli.mdx b/apps/web/src/content/docs/docs/v4.42.4/evaluation/batch-cli.mdx
index f9b204b46..8f8038e9a 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/evaluation/batch-cli.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/evaluation/batch-cli.mdx
@@ -238,24 +238,25 @@ expected_output:
 
 The grader extracts these fields and compares them against the parsed candidate output.
 
-## Target Configuration
+## Provider Configuration
 
-Configure the batch CLI provider in your targets file or eval file:
+Configure the batch CLI provider in your providers file or eval file:
 
 ```yaml
-# In agentv-targets.yaml or eval file
-targets:
-  batch_cli:
-    provider: cli
-    command: bun run ./scripts/batch-runner.ts --eval {EVAL_FILE} --output {OUTPUT_FILE}
-    provider_batching: true
+# In .agentv/providers.yaml or an eval file
+providers:
+  - id: cli
+    label: batch-cli
+    config:
+      command: bun run ./scripts/batch-runner.ts --eval {EVAL_FILE} --output {OUTPUT_FILE}
+      provider_batching: true
 ```
 
 Key settings:
 
 | Setting | Description |
 |---------|-------------|
-| `provider: cli` | Use the CLI provider |
+| `id: cli` | Use the CLI provider |
 | `provider_batching: true` | Run once for all tests instead of per-test |
 | `{EVAL_FILE}` | Placeholder replaced with the eval file path |
 | `{OUTPUT_FILE}` | Placeholder replaced with the JSONL output path |
diff --git a/apps/web/src/content/docs/docs/v4.42.4/evaluation/eval-files.mdx b/apps/web/src/content/docs/docs/v4.42.4/evaluation/eval-files.mdx
index 46cd3669c..405981fd6 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/evaluation/eval-files.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/evaluation/eval-files.mdx
@@ -174,14 +174,14 @@ prompts:
     label: Terse
     prompt: "In one sentence, summarize {{ vars.topic }}."
 
-targets:
-  - id: local-mini
-    provider: openai
+providers:
+  - id: openai
+    label: local-mini
     runtime: host
     config:
       model: gpt-5.4-mini
-  - id: local-codex
-    provider: codex-app-server
+  - id: codex-app-server
+    label: local-codex
     runtime: host
     config:
       command: ["codex", "app-server"]
diff --git a/apps/web/src/content/docs/docs/v4.42.4/evaluation/running-evals.mdx b/apps/web/src/content/docs/docs/v4.42.4/evaluation/running-evals.mdx
index 35fb1ee67..8a9fa9838 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/evaluation/running-evals.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/evaluation/running-evals.mdx
@@ -54,7 +54,7 @@ The `duration_ms`, `started_at`, and `ended_at` fields are present on every grad
 Run against a different target than specified in the eval file:
 
 ```bash
-agentv eval --target my-target evals/**/*.yaml
+agentv eval --provider my-target evals/**/*.yaml
 ```
 
 ### Experiment Label
@@ -215,7 +215,7 @@ agentv eval evals/my-eval.yaml --workers 4
 agentv eval evals/file1.yaml evals/file2.yaml evals/file3.yaml --workers 3
 # Files run one at a time; within each file, up to 3 test cases run in parallel
 
-agentv eval evals/my-eval.yaml --target gpt --target claude --workers 4
+agentv eval evals/my-eval.yaml --provider gpt --provider claude --workers 4
 # The target matrix shares the same 4-worker budget
 ```
 
@@ -462,22 +462,22 @@ Use `config.yaml` for portable defaults and shared eval-definition fields that c
 
 ### YAML config (`config.yaml` plus optional `config.local.yaml`)
 
-Project config and eval YAML share the same eval-definition graph for targets,
+Project config and eval YAML share the same eval-definition graph for providers,
 graders, tests, defaults, and execution policy. Small projects can keep that
 graph inline:
 
 ```yaml
-targets:
-  - id: codex-local
-    provider: codex-app-server
+providers:
+  - id: codex-app-server
+    label: codex-local
     runtime: host
     config:
       command: ["codex", "app-server"]
       model: gpt-5-codex
 
 graders:
-  - id: openai-grader
-    provider: openai
+  - id: openai
+    label: openai-grader
     config:
       model: gpt-5-mini
 
@@ -486,7 +486,7 @@ tests:
     input: Fix the failing test
 
 defaults:
-  target: codex-local
+  provider: codex-local
   grader: openai-grader
 
 execution:
@@ -498,7 +498,7 @@ Larger projects can decompose any supported top-level field with a direct
 directly:
 
 ```yaml
-targets: file://targets.yaml
+providers: file://providers.yaml
 graders: file://graders.yaml
 tests: file://tests.yaml
 defaults: file://defaults.yaml
@@ -506,9 +506,9 @@ execution: file://execution.yaml
 ```
 
 ```yaml
-# .agentv/targets.yaml
-- id: codex-local
-  provider: codex-app-server
+# .agentv/providers.yaml
+- id: codex-app-server
+  label: codex-local
   runtime: host
   config:
     command: ["codex", "app-server"]
@@ -516,13 +516,13 @@ execution: file://execution.yaml
 
 ```yaml
 # .agentv/defaults.yaml
-target: codex-local
+provider: codex-local
 grader: openai-grader
 ```
 
 Do not wrap referenced field files in another object. For example,
-`targets: file://targets.yaml` expects `targets.yaml` to contain a bare array,
-not `{ targets: [...] }`.
+`providers: file://providers.yaml` expects `providers.yaml` to contain a bare
+array, not `{ providers: [...] }`.
 
 In authored eval YAML, use `evaluate_options.max_concurrency` for suite
 parallelism. In `.agentv/config.yaml`, operators can set
@@ -598,31 +598,31 @@ export default defineConfig({
 
 Response cache and replay are separate concepts. The response cache is an iteration aid for repeated live provider calls. Transcript or fixture replay is target substitution from curated artifacts, and graders still run fresh against the replayed output.
 
-## Replay Target Fixtures
+## Replay Provider Fixtures
 
-Replay target fixtures let you record live target output once, then swap in a replay target alias for later runs without changing eval YAML or grader config. This is useful for expensive coding-agent and document-intelligence runs where you want deterministic target output but fresh grading.
+Replay provider fixtures let you record live provider output once, then swap in a replay provider alias for later runs without changing eval YAML or grader config. This is useful for expensive coding-agent and document-intelligence runs where you want deterministic provider output but fresh grading.
 
-Record target output from the live target with `--record-replay`:
+Record provider output from the live provider with `--record-replay`:
 
 ```bash
 agentv eval evals/legal-review.eval.yaml \
-  --target live_coding_agent \
+  --provider live_coding_agent \
   --record-replay fixtures/legal-review-target-output.jsonl
 ```
 
-Then add a replay target alias in `.agentv/targets.yaml`:
+Then add a replay provider alias in `.agentv/providers.yaml`:
 
 ```yaml
-targets:
-  - id: live_coding_agent
-    provider: codex-cli
+providers:
+  - id: codex-cli
+    label: live_coding_agent
     runtime: host
     config:
       command: ["codex"]
       model: gpt-5
 
-  - id: replay_coding_agent
-    provider: replay
+  - id: replay
+    label: replay_coding_agent
     runtime: host
     config:
       fixtures: ../fixtures/legal-review-target-output.jsonl
@@ -633,7 +633,7 @@ targets:
 Run the same eval against the replay alias:
 
 ```bash
-agentv eval evals/legal-review.eval.yaml --target replay_coding_agent
+agentv eval evals/legal-review.eval.yaml --provider replay_coding_agent
 ```
 
 Replay fixture rows are strict snake_case JSONL. Each row is keyed by `suite` or `eval_path`, `test_id`, `source_target`, `attempt`, and optional `variant`; missing or duplicate rows fail before grading. Rows preserve the recorded target `output`, `tool_calls`, `transcript`, `token_usage`, `cost_usd`, `duration_ms`, `start_time`, and `end_time` when the live provider supplied them.
diff --git a/apps/web/src/content/docs/docs/v4.42.4/evaluation/sdk.mdx b/apps/web/src/content/docs/docs/v4.42.4/evaluation/sdk.mdx
index 1186478b3..28df620b5 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/evaluation/sdk.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/evaluation/sdk.mdx
@@ -328,7 +328,7 @@ const { results, summary } = await evaluate({
 console.log(`${summary.passed}/${summary.total} passed`);
 ```
 
-Auto-discovers the `default` target from `.agentv/targets.yaml` and `.env` credentials.
+Auto-discovers the `default` provider from `.agentv/providers.yaml` and `.env` credentials.
 
 ### File-Based via `specFile`
 
diff --git a/apps/web/src/content/docs/docs/v4.42.4/getting-started/installation.mdx b/apps/web/src/content/docs/docs/v4.42.4/getting-started/installation.mdx
index e128d6bf2..dbc7ca6bf 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/getting-started/installation.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/getting-started/installation.mdx
@@ -46,13 +46,13 @@ agentv skills get --all                    # load all skills
 
 After setup, you should have:
 - `.agentv/config.yaml`
-- `.agentv/targets.yaml`
+- `.agentv/providers.yaml`
 - `.env.example`
 
 ```bash
 test -f .env.example
 test -f .agentv/config.yaml
-test -f .agentv/targets.yaml
+test -f .agentv/providers.yaml
 ```
 
 ## Claude Code Plugin (Optional)
diff --git a/apps/web/src/content/docs/docs/v4.42.4/getting-started/quickstart.mdx b/apps/web/src/content/docs/docs/v4.42.4/getting-started/quickstart.mdx
index fcba3976f..185adbb5f 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/getting-started/quickstart.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/getting-started/quickstart.mdx
@@ -37,7 +37,7 @@ development.
 
 1. Copy `.env.example` to `.env`
 2. Fill in your API keys, endpoints, and other configuration values
-3. Update the environment variable names in `.agentv/targets.yaml` to match the variables you
+3. Update the environment variable names in `.agentv/providers.yaml` to match the variables you
    exported or defined in `.env`
 
 ## 4. Create an eval
diff --git a/apps/web/src/content/docs/docs/v4.42.4/guides/skill-improvement-workflow.mdx b/apps/web/src/content/docs/docs/v4.42.4/guides/skill-improvement-workflow.mdx
index 7c14ebeb6..d69f5dfd6 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/guides/skill-improvement-workflow.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/guides/skill-improvement-workflow.mdx
@@ -93,7 +93,7 @@ Start with 5–10 focused test cases. You can always add more as you discover ed
 Run the evaluation **without** the skill loaded to establish a baseline:
 
 ```bash
-agentv eval evals.json --target baseline
+agentv eval evals.json --provider baseline
 ```
 
 This produces a results file (e.g., `results-baseline.jsonl`) showing how the agent performs on its own.
@@ -111,7 +111,7 @@ drafts/
     SKILL.md
 
 # Baseline run won't pick it up
-agentv eval evals.json --target baseline
+agentv eval evals.json --provider baseline
 ```
 
 ## Step 3: Run Candidate Evaluation
@@ -119,7 +119,7 @@ agentv eval evals.json --target baseline
 Run the same evaluation **with** the skill loaded:
 
 ```bash
-agentv eval evals.json --target candidate
+agentv eval evals.json --provider candidate
 ```
 
 Or grade existing sessions offline (no API keys required):
@@ -198,7 +198,7 @@ Loop back to Step 3 with the improved skill:
 
 ```bash
 # Run the improved candidate
-agentv eval evals.json --target candidate
+agentv eval evals.json --provider candidate
 
 # Compare against the previous baseline
 agentv compare results-baseline.jsonl results-candidate.jsonl
@@ -256,11 +256,11 @@ If you've been using the Agent Skills skill-creator workflow, AgentV reads your
 | Skill-Creator | AgentV | Notes |
 |--------------|--------|-------|
 | `evals.json` | `agentv eval evals.json` | Direct — no conversion needed |
-| `claude -p "prompt"` | `agentv eval evals.json --target claude` | Same eval, richer engine |
+| `claude -p "prompt"` | `agentv eval evals.json --provider claude` | Same eval, richer engine |
 | `grading.json` (read) | `<test-id>/grading.json` (write) | Same per-test schema, AgentV writes one grading file per test case |
 | `benchmark.json` (read) | `<output>/benchmark.json` (write) | AgentV writes the canonical run summary; convert it in a wrapper if another tool needs a narrower compatibility shape |
 | n/a | `index.jsonl` (write) | AgentV-specific per-test manifest for filtering, retry, and replay workflows |
-| with-skill vs without-skill | `--target baseline --target candidate` | Structured comparison |
+| with-skill vs without-skill | `--provider baseline --provider candidate` | Structured comparison |
 | Graduate to richer evals | `agentv convert evals.json` → EVAL.yaml | Adds workspace, code graders, etc. |
 
 **Key takeaway:** You do not need to rewrite your `evals.json`. AgentV reads it directly and adds a richer evaluation engine on top.
@@ -279,7 +279,7 @@ agentv pipeline run evals/my-eval.yaml --experiment with_skills
 
 Both runs use the same eval file and produce separate run directories. The experiment label is recorded in `manifest.json` and `index.jsonl`, making it easy to filter and compare in dashboards.
 
-This replaces the need for separate `--target baseline` / `--target candidate` configurations when the only difference between runs is the workspace setup (skills, config, etc.) rather than the target harness.
+This replaces the need for separate `--provider baseline` / `--provider candidate` configurations when the only difference between runs is the workspace setup (skills, config, etc.) rather than the provider harness.
 
 ## Baseline Comparison Best Practices
 
diff --git a/apps/web/src/content/docs/docs/v4.42.4/guides/workspace-architecture.mdx b/apps/web/src/content/docs/docs/v4.42.4/guides/workspace-architecture.mdx
index 13c9b487b..f79f4ff04 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/guides/workspace-architecture.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/guides/workspace-architecture.mdx
@@ -158,30 +158,30 @@ tested.
 ## World vs player boundary
 
 The eval workspace is the **world**: the same repos, fixtures, template files,
-and workspace hooks are shared by every target in the run. A target is the
+and workspace hooks are shared by every provider in the run. A provider is the
 **player**: the harness under evaluation, plus provider configuration and
-target-specific setup hooks.
+provider-specific setup hooks.
 
-Targets do **not** declare `repos`. Keeping repo provenance in the shared eval
-workspace is what makes multi-target comparison valid: every target sees the
+Providers do **not** declare `repos`. Keeping repo provenance in the shared eval
+workspace is what makes multi-provider comparison valid: every provider sees the
 same substrate, and differences in results come from the harness, not from a
 different checkout.
 
-Use target hooks for per-harness setup:
+Use provider hooks for per-harness setup:
 
 ```yaml
 execution:
-  targets:
+  providers:
     - baseline
     - name: with-skills
-      use_target: baseline
+      use_provider: baseline
       hooks:
         before_each:
           command: ["sh", "-c", "cp -R skills \"{{workspace_path}}/.claude/skills\""]
 ```
 
-Workspace hooks run first on setup, then target hooks. Teardown runs in the
-opposite order. See [Target Hooks](/docs/v4.42.4/targets/configuration/#target-hooks)
+Workspace hooks run first on setup, then provider hooks. Teardown runs in the
+opposite order. See [Provider Hooks](/docs/v4.42.4/targets/configuration/#provider-hooks)
 for the command schema and full lifecycle order.
 
 ## Windows performance guidance
diff --git a/apps/web/src/content/docs/docs/v4.42.4/integrations/agent-skills-evals.mdx b/apps/web/src/content/docs/docs/v4.42.4/integrations/agent-skills-evals.mdx
index 513cc30fc..bee673238 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/integrations/agent-skills-evals.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/integrations/agent-skills-evals.mdx
@@ -15,7 +15,7 @@ pagefind: false
 AgentV natively supports `evals.json`. You can run Agent Skills evals directly:
 
 ```bash
-agentv eval evals.json --target claude
+agentv eval evals.json --provider claude
 ```
 
 When you need AgentV's power features (deterministic graders, assert-set scoring, multi-turn conversations, workspace isolation), you can graduate to EVAL.yaml.
@@ -46,10 +46,10 @@ Create `evals.json`:
 Run it:
 
 ```bash
-agentv eval evals.json --target claude
+agentv eval evals.json --provider claude
 ```
 
-The `--target` flag selects the agent harness. The agent evaluates itself — skills load naturally via progressive disclosure.
+The `--provider` flag selects the agent harness. The agent evaluates itself — skills load naturally via progressive disclosure.
 
 ## Field mapping
 
@@ -113,7 +113,7 @@ The rest of the bundle follows the same pattern:
 Generate the run `benchmark.json` alongside the standard result JSONL. The `benchmark.json` is automatically written to the artifact directory:
 
 ```bash
-agentv eval evals.json --target claude --output ./results
+agentv eval evals.json --provider claude --output ./results
 # benchmark.json is written to ./results/benchmark.json
 ```
 
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/cli-provider.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/cli-provider.mdx
index 4d32842f6..669dcab85 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/cli-provider.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/cli-provider.mdx
@@ -1,6 +1,6 @@
 ---
 title: CLI Provider
-description: Wrap any shell command as an evaluation target
+description: Wrap any shell command as an evaluation provider
 sidebar:
   order: 4
 slug: docs/v4.42.4/targets/cli-provider
@@ -8,7 +8,7 @@ editUrl: false
 pagefind: false
 ---
 
-The `cli` provider runs an arbitrary shell command per test case and captures its output as the target's response. It's the escape hatch that lets you evaluate *anything* that exposes a command-line entry point — your own agent, a third-party CLI, a stub that prints a fixed answer, a script that calls an in-house microservice, etc.
+The `cli` provider runs an arbitrary shell command per test case and captures its output as the provider response. It's the escape hatch that lets you evaluate *anything* that exposes a command-line entry point — your own agent, a third-party CLI, a stub that prints a fixed answer, a script that calls an in-house microservice, etc.
 
 Because the contract is "we invoke a command and read a file," almost any useful composition pattern (sanity-checking your grader against a known-good answer, diffing two implementations, driving a batch mode) can be built on top without any new primitives.
 
@@ -16,23 +16,23 @@ Because the contract is "we invoke a command and read a file," almost any useful
 
 ```yaml
 # .agentv/config.yaml
-targets:
-  - id: my-agent
-    provider: cli
+providers:
+  - id: cli
+    label: my-agent
     runtime: host
     config:
       command: python agent.py --prompt {PROMPT} --out {OUTPUT_FILE}
 
 graders:
-  - id: azure-grader
-    provider: azure
+  - id: azure
+    label: azure-grader
     config:
       endpoint: ${{ AZURE_OPENAI_ENDPOINT }}
       api_key: ${{ AZURE_OPENAI_API_KEY }}
       model: ${{ AZURE_DEPLOYMENT_NAME }}
 
 defaults:
-  target: my-agent
+  provider: my-agent
   grader: azure-grader
 ```
 
@@ -45,7 +45,7 @@ Before each test case, AgentV renders the `command` template and spawns it as a
 1. **Read the input** via one of the placeholders below.
 2. **Write the response to `{OUTPUT_FILE}`** — AgentV reads *that file*, not your stdout.
 
-When the process exits successfully, AgentV parses the contents of `{OUTPUT_FILE}` and treats it as the target's response. Non-zero exits, timeouts, and unreadable output files are surfaced as test errors with the underlying stderr/exit code.
+When the process exits successfully, AgentV parses the contents of `{OUTPUT_FILE}` and treats it as the provider response. Non-zero exits, timeouts, and unreadable output files are surfaced as test errors with the underlying stderr/exit code.
 
 ### Template placeholders
 
@@ -86,9 +86,9 @@ echo "Hello, world!" > {OUTPUT_FILE}
 
 | Field | Type | Required | Default | Description |
 |---|---|---|---|---|
-| `id` | string | yes | — | AgentV target identity used by eval `target`, CLI `--target`, and comparisons. |
-| `provider` | literal `"cli"` | yes | — | Selects this provider. |
-| `runtime` | string or object | no | `host` | Target runtime placement. |
+| `id` | literal `"cli"` | yes | — | Selects the CLI backend/provider spec. |
+| `label` | string | yes | — | Stable AgentV selection and result identity used by eval `providers` and CLI `--provider`. |
+| `runtime` | string or object | no | `host` | Provider runtime placement. |
 | `config.command` | string | yes | — | Shell command template. |
 | `config.timeout_seconds` | number | no | — | Kill the process if it runs longer than this. |
 | `config.cwd` | string | no | eval dir | Working directory. Relative paths resolve against the eval file. |
@@ -100,12 +100,12 @@ echo "Hello, world!" > {OUTPUT_FILE}
 
 ## Batching
 
-For targets where spin-up cost dominates per-case work (e.g. loading a model, authenticating), set `batch_requests: true`. AgentV invokes the command *once*, hands it a JSONL stream of cases, and expects a JSONL response keyed by each case's `id`:
+For providers where spin-up cost dominates per-case work (e.g. loading a model, authenticating), set `batch_requests: true`. AgentV invokes the command *once*, hands it a JSONL stream of cases, and expects a JSONL response keyed by each case's `id`:
 
 ```yaml
-targets:
-  - id: batched-agent
-    provider: cli
+providers:
+  - id: cli
+    label: batched-agent
     runtime: host
     config:
       batch_requests: true
@@ -118,26 +118,26 @@ targets:
 
 A common question when building a new eval: **"if my grader scores my agent poorly, is the agent wrong or is the grader wrong?"** The classical testing answer is to run a known-correct reference ("the oracle") through the same grader — if a perfect answer doesn't pass, the grader is the bug.
 
-AgentV has no dedicated "oracle" feature because the `cli` provider already composes into one. Declare a second target that prints your known-good answer into `{OUTPUT_FILE}`, run the same eval against it, and assert a perfect score:
+AgentV has no dedicated "oracle" feature because the `cli` provider already composes into one. Declare a second provider that prints your known-good answer into `{OUTPUT_FILE}`, run the same eval against it, and assert a perfect score:
 
 ```yaml
 # .agentv/config.yaml
-targets:
-  - id: my-agent
-    provider: cli
+providers:
+  - id: cli
+    label: my-agent
     runtime: host
     config:
       command: python agent.py --prompt {PROMPT} --out {OUTPUT_FILE}
 
-  - id: oracle
-    provider: cli
+  - id: cli
+    label: oracle
     runtime: host
     config:
       command: cp fixtures/{EVAL_ID}.expected.txt {OUTPUT_FILE}
 
 graders:
-  - id: azure-grader
-    provider: azure
+  - id: azure
+    label: azure-grader
     config:
       endpoint: ${{ AZURE_OPENAI_ENDPOINT }}
       api_key: ${{ AZURE_OPENAI_API_KEY }}
@@ -150,29 +150,29 @@ defaults:
 ```bash
 # While iterating on your grader, run the oracle first.
 # If it doesn't score 100%, fix the grader before trusting any agent results.
-agentv eval my.EVAL.yaml --target oracle
+agentv eval my.EVAL.yaml --provider oracle
 
-# Then run the real target.
-agentv eval my.EVAL.yaml --target my-agent
+# Then run the real provider.
+agentv eval my.EVAL.yaml --provider my-agent
 ```
 
 A few practical notes:
 
-- `{EVAL_ID}` in the oracle command lets one target serve an entire eval suite — just ship one `fixtures/<id>.expected.txt` per case. Alternatively, read the expected output from wherever your rubric already keeps it.
+- `{EVAL_ID}` in the oracle command lets one provider serve an entire eval suite — just ship one `fixtures/<id>.expected.txt` per case. Alternatively, read the expected output from wherever your rubric already keeps it.
 - If the oracle doesn't reach 100%, that's the bug. Do not proceed to scoring real agents until it does.
 - If the oracle *does* reach 100%, low scores on real agents are a signal about the agent, not the grader.
-- The same composition works for other meta-tests: a "deliberately wrong" target that should score 0, a "mostly right" target pinned at a known partial score, etc.
+- The same composition works for other meta-tests: a "deliberately wrong" provider that should score 0, a "mostly right" provider pinned at a known partial score, etc.
 
-The pattern needs no special config field, no directory convention, and no flag — it's just a second target that happens to know the answer.
+The pattern needs no special config field, no directory convention, and no flag — it's just a second provider that happens to know the answer.
 
 Use this pattern instead of eval mock dry-run for grader validation. Mock
 execution was removed because fake candidate answers produced misleading
-quality failures; a reference target gives deterministic output while exercising
+quality failures; a reference provider gives deterministic output while exercising
 the real grader path.
 
 ## Debugging
 
-When a `cli` target misbehaves:
+When a `cli` provider misbehaves:
 
 1. Set `verbose: true` to see the rendered command and cwd.
 2. Set `keep_temp_files: true` and inspect `{PROMPT_FILE}` / `{OUTPUT_FILE}` after the run.
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/coding-agents.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/coding-agents.mdx
index 00e625ab7..5e823362a 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/coding-agents.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/coding-agents.mdx
@@ -1,6 +1,6 @@
 ---
 title: Coding Agents
-description: Evaluate coding agent targets
+description: Evaluate coding agent providers
 sidebar:
   order: 3
 slug: docs/v4.42.4/targets/coding-agents
@@ -8,21 +8,21 @@ editUrl: false
 pagefind: false
 ---
 
-Coding agent targets evaluate assistants that run against a real workspace. They
-use the target shape `id`, `provider`, `runtime`, and `config`. The target `id`
-is AgentV's stable selection and artifact identity. `provider` names the
-adapter or control boundary. `runtime` names where the target runs. Provider
-settings live under `config`.
+Coding agent providers evaluate assistants that run against a real workspace.
+They use the provider shape `id`, `label`, `runtime`, and `config`. `id` names
+the backend or package provider spec. `label` is AgentV's stable selection and
+artifact identity. `runtime` names where the provider runs. Provider settings
+live under `config`.
 
-Use `defaults.grader`, CLI `--grader` / `--grader-target`, or an
-evaluator-specific target override for LLM-based grading. Grader selection is
-separate from the coding-agent target, so target definitions do not carry a
+Use `defaults.grader`, CLI `--grader` / `--grader-provider`, or an
+evaluator-specific provider override for LLM-based grading. Grader selection is
+separate from the coding-agent provider, so provider definitions do not carry a
 grader field.
 
 ```yaml
-targets:
-  - id: codex-local
-    provider: codex-app-server
+providers:
+  - id: codex-app-server
+    label: codex-local
     runtime: host
     config:
       command: ["codex", "app-server"]
@@ -30,13 +30,13 @@ targets:
       reasoning_effort: high
 
 graders:
-  - id: openai-grader
-    provider: openai
+  - id: openai
+    label: openai-grader
     config:
       model: gpt-5-mini
 
 defaults:
-  target: codex-local
+  provider: codex-local
   grader: openai-grader
 
 execution:
@@ -61,9 +61,9 @@ be evaluated only by intentionally mounting or seeding the profile directory
 the agent needs, which trades portability for fidelity to the local agent setup.
 
 ```yaml
-targets:
-  - id: codex-clean-profile
-    provider: codex-cli
+providers:
+  - id: codex-cli
+    label: codex-clean-profile
     runtime:
       mode: profile
       codex_home: .agentv/profiles/codex-clean
@@ -76,9 +76,9 @@ targets:
 ```
 
 ```yaml
-targets:
-  - id: codex-ci-sandbox
-    provider: codex-cli
+providers:
+  - id: codex-cli
+    label: codex-ci-sandbox
     runtime:
       mode: sandbox
       engine: docker
@@ -137,7 +137,7 @@ bun add --optional @anthropic-ai/claude-agent-sdk
 bun add --optional @github/copilot-sdk
 ```
 
-Missing SDK packages are reported as target execution failures from the child
+Missing SDK packages are reported as provider execution failures from the child
 runner. Prefer the process/protocol providers for default local and CI runs,
 and choose an SDK provider only when SDK-native events or lifecycle control are
 the behavior under test.
@@ -147,9 +147,9 @@ the behavior under test.
 Use `codex-app-server` when you want rich protocol control:
 
 ```yaml
-targets:
-  - id: codex-local
-    provider: codex-app-server
+providers:
+  - id: codex-app-server
+    label: codex-local
     runtime: host
     config:
       command: ["codex", "app-server"]
@@ -162,9 +162,9 @@ Use `codex-cli` when a simple CLI boundary is enough or when you want an
 operator-specific shim:
 
 ```yaml
-targets:
-  - id: codex-eng
-    provider: codex-cli
+providers:
+  - id: codex-cli
+    label: codex-eng
     runtime: host
     config:
       command: ["codex-eng", "exec", "--json"]
@@ -174,9 +174,9 @@ targets:
 Use `codex-sdk` only when you intentionally want the Codex SDK path:
 
 ```yaml
-targets:
-  - id: codex-sdk-isolated
-    provider: codex-sdk
+providers:
+  - id: codex-sdk
+    label: codex-sdk-isolated
     runtime: host
     config:
       model: gpt-5-codex
@@ -192,9 +192,9 @@ Common Codex config fields include `command`, `model`, `reasoning_effort`,
 Use `pi-rpc` for the rich stdio/RPC boundary:
 
 ```yaml
-targets:
-  - id: pi-rpc-local
-    provider: pi-rpc
+providers:
+  - id: pi-rpc
+    label: pi-rpc-local
     runtime: host
     config:
       command: ["pi"]
@@ -205,9 +205,9 @@ targets:
 Use `pi-cli` for simple subprocess execution:
 
 ```yaml
-targets:
-  - id: pi-cli-local
-    provider: pi-cli
+providers:
+  - id: pi-cli
+    label: pi-cli-local
     runtime:
       mode: profile
       home: .agentv/profiles/pi-local
@@ -221,9 +221,9 @@ targets:
 Use `pi-sdk` only when you intentionally want the SDK path:
 
 ```yaml
-targets:
-  - id: pi-sdk-isolated
-    provider: pi-sdk
+providers:
+  - id: pi-sdk
+    label: pi-sdk-isolated
     runtime: host
     config:
       subprovider: openai-codex
@@ -240,9 +240,9 @@ Azure provider path for custom gateways.
 ## Claude
 
 ```yaml
-targets:
-  - id: claude-local
-    provider: claude-cli
+providers:
+  - id: claude-cli
+    label: claude-local
     runtime: host
     config:
       command: ["claude"]
@@ -255,9 +255,9 @@ locally. Use `claude-sdk` only when you intentionally want the Claude Agent SDK
 path:
 
 ```yaml
-targets:
-  - id: claude-sdk-isolated
-    provider: claude-sdk
+providers:
+  - id: claude-sdk
+    label: claude-sdk-isolated
     runtime: host
     config:
       model: claude-sonnet-4-20250514
@@ -271,9 +271,9 @@ and `system_prompt`.
 ## Copilot
 
 ```yaml
-targets:
-  - id: copilot-local
-    provider: copilot-cli
+providers:
+  - id: copilot-cli
+    label: copilot-local
     runtime: host
     config:
       command: ["copilot"]
@@ -283,9 +283,9 @@ targets:
 Route Copilot through an OpenAI-compatible endpoint:
 
 ```yaml
-targets:
-  - id: copilot-openai
-    provider: copilot-cli
+providers:
+  - id: copilot-cli
+    label: copilot-openai
     runtime: host
     config:
       command: ["copilot"]
@@ -303,9 +303,9 @@ agentv import copilot --session-id <uuid> -o .agentv/transcripts/copilot-case.js
 ```
 
 ```yaml
-targets:
-  - id: copilot-session-replay
-    provider: replay
+providers:
+  - id: replay
+    label: copilot-session-replay
     config:
       transcripts: .agentv/transcripts/copilot-case.jsonl
       source_target: copilot-cli
@@ -314,9 +314,9 @@ targets:
 Use `copilot-sdk` only when you intentionally want the SDK path:
 
 ```yaml
-targets:
-  - id: copilot-sdk-isolated
-    provider: copilot-sdk
+providers:
+  - id: copilot-sdk
+    label: copilot-sdk-isolated
     runtime: host
     config:
       model: gpt-5-mini
@@ -366,9 +366,9 @@ LLM providers receive file content inline instead; see
 For deterministic harness checks without a real provider:
 
 ```yaml
-targets:
-  - id: mock-target
-    provider: mock
+providers:
+  - id: mock
+    label: mock-target
     runtime: host
     config:
       response: ok
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/configuration.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/configuration.mdx
index 251e0f516..b1f24b37a 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/configuration.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/configuration.mdx
@@ -1,6 +1,6 @@
 ---
-title: Targets Configuration
-description: Configure execution targets for providers and agents
+title: Providers Configuration
+description: Configure candidate, grader, and agent providers
 sidebar:
   order: 1
 slug: docs/v4.42.4/targets/configuration
@@ -8,29 +8,29 @@ editUrl: false
 pagefind: false
 ---
 
-Targets define which agent or LLM provider to evaluate. AgentV uses one
+Providers define which agent or LLM provider to evaluate. AgentV uses one
 composable config graph across project manifests and eval files:
 
 - `.agentv/config.yaml` is the project-local discovery and composition root. It
-  can hold targets, graders, tests, defaults, execution policy, results
+  can hold providers, graders, tests, defaults, execution policy, results
   settings, and repo-local project policy.
 - `$AGENTV_HOME/config.yaml` is the user/operator config. Use it for defaults
   that apply across projects, project registry data, default result locations,
   and provider defaults that should not be copied into each repo.
 - `eval.yaml` is a focused, shareable slice of the same graph. Use it for a
-  suite-specific target, grader, tests, evaluator settings, or run controls
+  suite-specific provider, grader, tests, evaluator settings, or run controls
   that should travel with the eval.
 
 Any supported top-level field can stay inline or become a direct field
-reference such as `targets: file://targets.yaml`. Both forms normalize to the
+reference such as `providers: file://providers.yaml`. Both forms normalize to the
 same config graph.
 
 ## Structure
 
 ```yaml
-targets:
-  - id: local-openai
-    provider: openai
+providers:
+  - id: openai
+    label: local-openai
     runtime: host
     config:
       api_format: chat
@@ -38,52 +38,53 @@ targets:
       api_key: ${{ LOCAL_OPENAI_PROXY_API_KEY }}
       model: ${{ LOCAL_OPENAI_PROXY_MODEL }}
 
-  - id: codex-local
-    provider: codex-app-server
+  - id: codex-app-server
+    label: codex-local
     runtime: host
     config:
       command: ["codex", "app-server"]
       model: gpt-5-codex
 
 graders:
-  - id: openai-grader
-    provider: openai
+  - id: openai
+    label: openai-grader
     config:
       model: gpt-5-mini
 
 defaults:
-  target: codex-local
+  provider: codex-local
   grader: openai-grader
 ```
 
-Use `id` for the stable AgentV target identity. `provider` selects the adapter
-or control boundary. `runtime` describes where the provider runs; use `host` as
-the shorthand for the current machine, or object form when you need
+Use `id` for the Promptfoo-compatible backend or package provider spec. Use
+`label` for the stable AgentV selection and result identity. `runtime` describes
+where the provider runs; use `host` as the shorthand for the current machine, or
+object form when you need
 `mode: host | profile | sandbox` plus runtime-specific settings. Provider
 settings belong under `config`. Process-backed coding-agent providers use
 `config.command` as a non-empty argv array.
 
 ## Runtime Modes
 
-Use `runtime: host` when you want AgentV to run the target exactly as it is
+Use `runtime: host` when you want AgentV to run the provider exactly as it is
 installed on the current machine. This is the best fit for local research,
 subscription-auth workflows, and evaluating the same CLI profile an engineer
 uses manually.
 
-Use `runtime.mode: profile` when the target still runs as a host process but
+Use `runtime.mode: profile` when the provider still runs as a host process but
 should use an isolated home/config directory, such as a dedicated `CODEX_HOME`
 or `HOME`.
 
-Use `runtime.mode: sandbox` when the target should run inside a separate
+Use `runtime.mode: sandbox` when the provider should run inside a separate
 execution substrate. The built-in sandbox runner currently supports Docker for
-`provider: cli`; provider-specific coding-agent adapters such as `codex-cli`,
-`claude-cli`, `copilot-cli`, and `pi-cli` return a structured unsupported target
+`id: cli`; provider-specific coding-agent adapters such as `codex-cli`,
+`claude-cli`, `copilot-cli`, and `pi-cli` return a structured unsupported provider
 error until their transcript parsers are wired through sandbox-aware runners.
 
 ```yaml
-targets:
-  - id: codex-sandbox
-    provider: codex-cli
+providers:
+  - id: codex-cli
+    label: codex-sandbox
     runtime:
       mode: sandbox
       engine: docker
@@ -107,9 +108,9 @@ targets:
 ```
 
 Sandbox mode does not inherit host credentials by default. Mount only the
-workspace, results, cache, or credential paths the target needs, and pass only
+workspace, results, cache, or credential paths the provider needs, and pass only
 the environment variables and secrets listed under `runtime.env` and
-`runtime.secrets`. Install the target CLI by using an image that already
+`runtime.secrets`. Install the provider CLI by using an image that already
 contains it or by adding explicit setup under `runtime.setup`; locate the CLI
 with `config.command`.
 
@@ -122,17 +123,17 @@ local subscription profile is the point of the evaluation.
 Inline and decomposed forms are equivalent. This single-file config:
 
 ```yaml
-targets:
-  - id: codex-local
-    provider: codex-app-server
+providers:
+  - id: codex-app-server
+    label: codex-local
     runtime: host
     config:
       command: ["codex", "app-server"]
       model: gpt-5-codex
 
 graders:
-  - id: openai-grader
-    provider: openai
+  - id: openai
+    label: openai-grader
     config:
       model: gpt-5-mini
 
@@ -141,26 +142,26 @@ tests:
     input: Fix the failing test.
 
 defaults:
-  target: codex-local
+  provider: codex-local
   grader: openai-grader
 ```
 
 can be decomposed like this:
 
 ```yaml
-targets: file://targets.yaml
+providers: file://providers.yaml
 graders: file://graders.yaml
 tests: file://tests.yaml
 defaults: file://defaults.yaml
 ```
 
-Referenced field files contain the field value directly. `targets.yaml` contains
-a bare array, not an object wrapped in `targets:`:
+Referenced field files contain the field value directly. `providers.yaml` contains
+a bare array, not an object wrapped in `providers:`:
 
 ```yaml
-# .agentv/targets.yaml
-- id: codex-local
-  provider: codex-app-server
+# .agentv/providers.yaml
+- id: codex-app-server
+  label: codex-local
   runtime: host
   config:
     command: ["codex", "app-server"]
@@ -169,7 +170,7 @@ a bare array, not an object wrapped in `targets:`:
 
 ```yaml
 # .agentv/defaults.yaml
-target: codex-local
+provider: codex-local
 grader: openai-grader
 ```
 
@@ -183,9 +184,9 @@ exported process environment variables directly, and it also loads `.env` files
 eval directory hierarchy when present:
 
 ```yaml
-targets:
-  - id: my-target
-    provider: anthropic
+providers:
+  - id: anthropic
+    label: my-provider
     runtime: host
     config:
       api_key: ${{ ANTHROPIC_API_KEY }}
@@ -215,39 +216,40 @@ already-exported secrets into `.env`.
 | `vscode` | Agent | VS Code with Copilot |
 | `vscode-insiders` | Agent | VS Code Insiders |
 | `cli` | Agent | Any CLI command — see [CLI Provider](/docs/targets/cli-provider) |
-| `mock` | Testing | Explicit mock target for examples and tests |
+| `mock` | Testing | Explicit mock provider for examples and tests |
 
-## Referencing Targets in Evals
+## Referencing Providers in Evals
 
-Select the system under test with `defaults.target`, top-level `target`, or CLI
-`--target`, depending on the command flow. Test cases do not choose targets;
-split target-specific cases into separate eval suites, select them with
-tags/filters, or run the same eval with different `--target` values.
+Select the system under test with `defaults.provider`, top-level `providers`, or
+CLI `--provider`, depending on the command flow. Test cases do not choose
+providers; split provider-specific cases into separate eval suites, select them
+with tags/filters, or run the same eval with different `--provider` values.
 
 ```yaml
-target: local-openai
+providers:
+  - local-openai
 
 tests:
   - id: test-1
   - id: test-2
 ```
 
-The string is a configured target `id`. Use object form when an eval needs a
-local target variant:
+The string is a configured provider `label`. Use object form when an eval needs
+a local provider variant:
 
 ```yaml
-target:
-  id: codex-high-reasoning
-  provider: codex-app-server
-  runtime: host
-  config:
-    command: ["codex", "app-server"]
-    model: gpt-5-codex
-    reasoning_effort: high
+providers:
+  - id: codex-app-server
+    label: codex-high-reasoning
+    runtime: host
+    config:
+      command: ["codex", "app-server"]
+      model: gpt-5-codex
+      reasoning_effort: high
 ```
 
 Use `defaults.grader` for the project default grader. A specific evaluator can
-still choose its own grader target when the evaluator supports that override.
+still choose its own grader provider when the evaluator supports that override.
 
 ### Lifecycle Extensions
 
@@ -255,7 +257,7 @@ Run non-provisioning setup at Promptfoo-compatible lifecycle points using
 top-level `extensions`. The harness materializes `workspace.template` and
 `workspace.repos` first, then runs `beforeAll` extensions. Use extensions for
 dependency installs, builds, fixture generation, and agent-rule staging. Use
-target hooks for runner-specific setup. Keep repo identity and checkout pins in
+provider hooks for runner-specific setup. Keep repo identity and checkout pins in
 `workspace.repos`; extensions must not become the default repo acquisition path.
 
 ```yaml
@@ -282,7 +284,7 @@ workspace:
 | `extensions[]` | `file://...:beforeAll`, `beforeEach`, `afterEach`, `afterAll`, or `agentv:agent-rules` |
 | `hooks.after_each.reset` | Reset mode: `none`, `fast`, `strict` |
 
-**Lifecycle order:** template copy → repo materialization → `extensions.beforeAll` → target `hooks.before_all` → git baseline → (`extensions.beforeEach` → target `hooks.before_each` → agent runs → file changes captured → target `hooks.after_each` → `extensions.afterEach` → `workspace.hooks.after_each.reset`) × N tests → target `hooks.after_all` → `extensions.afterAll` → cleanup
+**Lifecycle order:** template copy → repo materialization → `extensions.beforeAll` → provider `hooks.before_all` → git baseline → (`extensions.beforeEach` → provider `hooks.before_each` → agent runs → file changes captured → provider `hooks.after_each` → `extensions.afterEach` → `workspace.hooks.after_each.reset`) × N tests → provider `hooks.after_all` → `extensions.afterAll` → cleanup
 
 **Shared workspace:** The workspace is created once and shared across all tests in a suite. Use `hooks.after_each.reset` to reset state between tests (e.g., `fast`/`strict`).
 
@@ -337,7 +339,7 @@ workspace:
 | `scope` | `suite` reuses one harness-managed workspace for the suite; `attempt` creates a clean workspace for each resolved execution attempt |
 | `hooks.enabled` | Boolean (default: `true`). Set `false` to skip all lifecycle hooks. |
 
-Use `scope: attempt` when mutating agents need clean filesystem state for every prompt-target-test-repeat execution. Use `scope: suite` when the suite intentionally shares state across tests.
+Use `scope: attempt` when mutating agents need clean filesystem state for every prompt-provider-test-repeat execution. Use `scope: suite` when the suite intentionally shares state across tests.
 
 **Existing local workspaces:** do not commit local paths in eval YAML. Use `--workspace-path /path/to/workspace` for a one-off run, or put `execution.workspace_path` in `.agentv/config.local.yaml`.
 
@@ -385,42 +387,43 @@ CLI overrides:
 
 ### cwd
 
-Use `cwd` on a target to run in an existing directory (shared across tests). If not set, the eval file's directory is used as the working directory.
+Use `cwd` on a provider to run in an existing directory (shared across tests). If not set, the eval file's directory is used as the working directory.
 
-## Target Hooks
+## Provider Hooks
 
-Eval files can define per-target hooks that run setup/teardown scripts to customize the workspace for each target variant. This enables comparing different harness configurations (e.g., baseline vs with-plugins) in a single eval file.
+Eval files can define per-provider hooks that run setup/teardown scripts to customize the workspace for each provider variant. This enables comparing different harness configurations (e.g., baseline vs with-plugins) in a single eval file.
 
-Targets do not declare `repos`. Repositories belong to the shared eval workspace so every target runs in the same world; target hooks customize the harness under evaluation. Use hooks for per-target setup such as enabling wrappers or changing provider-local config. Keep installs, builds, fixture generation, and case setup in top-level lifecycle `extensions`.
+Providers do not declare `repos`. Repositories belong to the shared eval workspace so every provider runs in the same world; provider hooks customize the harness under evaluation. Use hooks for per-provider setup such as enabling wrappers or changing provider-local config. Keep installs, builds, fixture generation, and case setup in top-level lifecycle `extensions`.
 
-Target hooks can be scoped to an eval-local target object:
+Provider hooks can be scoped to an eval-local provider object:
 
 ```yaml
-target:
-  extends: default
-  hooks:
-    before_each:
-      command: ["setup-plugins.sh", "skills"]
+providers:
+  - label: candidate-with-skills
+    extends: default
+    hooks:
+      before_each:
+        command: ["setup-plugins.sh", "skills"]
 ```
 
 ### Hook execution order
 
-Target hooks run after workspace hooks on setup, before workspace hooks on teardown:
+Provider hooks run after workspace hooks on setup, before workspace hooks on teardown:
 
 1. Extension `beforeAll`
-2. **Target `before_all`**
+2. **Provider `before_all`**
 3. For each test:
    - Workspace `before_each`
-   - **Target `before_each`**
+   - **Provider `before_each`**
    - Test executes
-   - **Target `after_each`**
+   - **Provider `after_each`**
    - Workspace `after_each`
-4. **Target `after_all`**
+4. **Provider `after_all`**
 5. Workspace `after_all`
 
 ### Hook schema
 
-Target hooks follow the same schema as workspace hooks:
+Provider hooks follow the same schema as workspace hooks:
 
 ```yaml
 hooks:
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/custom-providers.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/custom-providers.mdx
index 40a6971bd..4455a220f 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/custom-providers.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/custom-providers.mdx
@@ -8,7 +8,7 @@ editUrl: false
 pagefind: false
 ---
 
-Custom providers let you implement evaluation targets in TypeScript instead of shelling out to a CLI command. This is useful when you want to call an HTTP API, use an SDK, or implement custom logic that goes beyond what the CLI provider supports.
+Custom providers let you implement evaluation providers in TypeScript instead of shelling out to a CLI command. This is useful when you want to call an HTTP API, use an SDK, or implement custom logic that goes beyond what the CLI provider supports.
 
 ## Provider Interface
 
@@ -175,20 +175,20 @@ registry.register('http-agent', (target: ResolvedTarget) => {
 });
 ```
 
-Then reference it in your targets file:
+Then reference it in your providers file:
 
 ```yaml
 # .agentv/config.yaml
-targets:
-  - id: my-http-agent
-    provider: http-agent
+providers:
+  - id: http-agent
+    label: my-http-agent
     runtime: host
     config:
       base_url: http://localhost:8080
 
 graders:
-  - id: azure-grader
-    provider: azure
+  - id: azure
+    label: azure-grader
     config:
       endpoint: ${{ AZURE_OPENAI_ENDPOINT }}
       api_key: ${{ AZURE_OPENAI_API_KEY }}
@@ -204,11 +204,11 @@ Custom provider kinds are not validated against the built-in provider list. When
 
 ## CLI Providers vs Native Providers
 
-AgentV supports two approaches for custom targets:
+AgentV supports two approaches for custom providers:
 
 | Aspect | CLI Provider | Native TypeScript Provider |
 |--------|-------------|---------------------------|
-| **Configuration** | YAML only (`provider: cli`) | TypeScript code + YAML |
+| **Configuration** | YAML only (`id: cli`) | TypeScript code + YAML |
 | **Communication** | Shell command + JSON output file | Direct function call |
 | **Best for** | Wrapping existing scripts, polyglot tools | HTTP APIs, SDKs, complex orchestration |
 | **Setup** | No code required | Requires a TypeScript entry point |
@@ -217,15 +217,15 @@ AgentV supports two approaches for custom targets:
 
 ### When to use CLI providers
 
-Use `provider: cli` when:
+Use `id: cli` when:
 - You have an existing script or binary to wrap
 - The agent is written in a different language (Python, Go, etc.)
 - You want zero TypeScript code
 
 ```yaml
-targets:
-  - id: python-agent
-    provider: cli
+providers:
+  - id: cli
+    label: python-agent
     runtime: host
     config:
       command: 'python agent.py --prompt-file {PROMPT_FILE} --output {OUTPUT_FILE}'
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/llm-providers.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/llm-providers.mdx
index 53dcfdae5..3ce7d5469 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/llm-providers.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/llm-providers.mdx
@@ -1,6 +1,6 @@
 ---
 title: LLM Providers
-description: Direct LLM API provider targets
+description: Direct LLM API providers
 sidebar:
   order: 2
 slug: docs/v4.42.4/targets/llm-providers
@@ -8,14 +8,14 @@ editUrl: false
 pagefind: false
 ---
 
-LLM provider targets call language model APIs directly. These are used both as evaluation targets and as grader targets for scoring.
+LLM providers call language model APIs directly. These are used both as candidate providers and as grader providers for scoring.
 
 ## OpenAI
 
 ```yaml
-targets:
-  - id: openai-target
-    provider: openai
+providers:
+  - id: openai
+    label: openai-target
     runtime: host
     config:
       api_key: ${{ OPENAI_API_KEY }}
@@ -42,9 +42,9 @@ Most users should leave this unset. The default `chat` format is universally sup
 
 ```yaml
 # OpenAI-compatible endpoint (default chat format works)
-targets:
-  - id: github-models
-    provider: openai
+providers:
+  - id: openai
+    label: github-models
     runtime: host
     config:
       api_format: chat
@@ -53,8 +53,8 @@ targets:
       model: ${{ GH_MODELS_MODEL }}
 
   # Opt in to Responses API for api.openai.com
-  - id: openai-responses
-    provider: openai
+  - id: openai
+    label: openai-responses
     runtime: host
     config:
       api_format: responses
@@ -69,8 +69,8 @@ the endpoint, model, and placeholder key in environment variables:
 
 ```yaml
 graders:
-  - id: local-openai-grader
-    provider: openai
+  - id: openai
+    label: local-openai-grader
     config:
       api_format: chat
       base_url: ${{ LOCAL_OPENAI_PROXY_BASE_URL }}
@@ -86,9 +86,9 @@ shared eval files.
 ## Azure OpenAI
 
 ```yaml
-targets:
-  - id: azure-base
-    provider: azure
+providers:
+  - id: azure
+    label: azure-base
     runtime: host
     config:
       endpoint: ${{ AZURE_OPENAI_ENDPOINT }}
@@ -103,16 +103,16 @@ targets:
 | `model` | Yes | Deployment name |
 | `version` | No | Azure API version (defaults to `v1`) |
 
-Azure targets always route through the Responses API (`/openai/v1/responses`). The api version defaults to `v1` and can be overridden via the `version` field.
+Azure providers always route through the Responses API (`/openai/v1/responses`). The api version defaults to `v1` and can be overridden via the `version` field.
 
 ### Chat-completions-only deployments
 
-If your Azure deployment only exposes `/chat/completions` (older deployments, certain regions), use `provider: openai` with a deployment-scoped `base_url` instead:
+If your Azure deployment only exposes `/chat/completions` (older deployments, certain regions), use `id: openai` with a deployment-scoped `base_url` instead:
 
 ```yaml
-targets:
-  - id: azure-chat
-    provider: openai
+providers:
+  - id: openai
+    label: azure-chat
     runtime: host
     config:
       base_url: https://<resource>.openai.azure.com/openai/deployments/<deployment>
@@ -121,14 +121,14 @@ targets:
       api_format: chat
 ```
 
-The `api_format` field was previously available on `provider: azure` but has been removed — Azure targets always go through the Responses API.
+The `api_format` field was previously available on `id: azure` but has been removed; Azure providers always go through the Responses API.
 
 ## Anthropic
 
 ```yaml
-targets:
-  - id: claude-target
-    provider: anthropic
+providers:
+  - id: anthropic
+    label: claude-target
     runtime: host
     config:
       api_key: ${{ ANTHROPIC_API_KEY }}
@@ -143,9 +143,9 @@ targets:
 ## Google Gemini
 
 ```yaml
-targets:
-  - id: gemini-target
-    provider: gemini
+providers:
+  - id: gemini
+    label: gemini-target
     runtime: host
     config:
       api_key: ${{ GEMINI_API_KEY }}
diff --git a/apps/web/src/content/docs/docs/v4.42.4/targets/retry.mdx b/apps/web/src/content/docs/docs/v4.42.4/targets/retry.mdx
index bb63b0509..7d64826a6 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/targets/retry.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/targets/retry.mdx
@@ -15,9 +15,9 @@ Configure automatic retry with exponential backoff for transient failures.
 Add retry fields to any target:
 
 ```yaml
-targets:
-  - id: azure-base
-    provider: azure
+providers:
+  - id: azure
+    label: azure-base
     runtime: host
     config:
       endpoint: ${{ AZURE_OPENAI_ENDPOINT }}
diff --git a/apps/web/src/content/docs/docs/v4.42.4/tools/compare.mdx b/apps/web/src/content/docs/docs/v4.42.4/tools/compare.mdx
index 14d735493..18cd82bc4 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/tools/compare.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/tools/compare.mdx
@@ -126,10 +126,10 @@ Compare different model versions:
 
 ```bash
 # Run baseline evaluation
-agentv eval evals/*.yaml --target gpt-4 --output runs/baseline
+agentv eval evals/*.yaml --provider gpt-4 --output runs/baseline
 
 # Run candidate evaluation
-agentv eval evals/*.yaml --target gpt-4o --output runs/candidate
+agentv eval evals/*.yaml --provider gpt-4o --output runs/candidate
 
 # Compare results
 agentv compare runs/baseline/index.jsonl runs/candidate/index.jsonl
diff --git a/apps/web/src/content/docs/docs/v4.42.4/tools/dashboard.mdx b/apps/web/src/content/docs/docs/v4.42.4/tools/dashboard.mdx
index aafa46db4..e0a20fc40 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/tools/dashboard.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/tools/dashboard.mdx
@@ -131,10 +131,10 @@ The default view shows a cross-experiment, cross-target performance matrix. Numb
 Run the same eval against multiple providers or experiment variants, then open the Analytics tab:
 
 ```bash
-agentv eval my.EVAL.yaml --target azure --experiment baseline
-agentv eval my.EVAL.yaml --target azure --experiment with-caching
-agentv eval my.EVAL.yaml --target gemini --experiment baseline
-agentv eval my.EVAL.yaml --target gemini --experiment with-caching
+agentv eval my.EVAL.yaml --provider azure --experiment baseline
+agentv eval my.EVAL.yaml --provider azure --experiment with-caching
+agentv eval my.EVAL.yaml --provider gemini --experiment baseline
+agentv eval my.EVAL.yaml --provider gemini --experiment with-caching
 agentv dashboard  # Analytics tab shows 2x2 matrix
 ```
 
diff --git a/apps/web/src/content/docs/docs/v4.42.4/tools/import.mdx b/apps/web/src/content/docs/docs/v4.42.4/tools/import.mdx
index 165dfe6b0..63df559b9 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/tools/import.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/tools/import.mdx
@@ -216,9 +216,9 @@ agentv eval evals/my-eval.yaml --transcript .agentv/transcripts/claude-4c4f9e4e.
 You can also use the same normalized transcript rows through a replay target:
 
 ```yaml
-targets:
-  - id: recorded-agent
-    provider: replay
+providers:
+  - id: replay
+    label: recorded-agent
     transcripts: .agentv/transcripts/claude-4c4f9e4e.jsonl
     source_target: claude-cli
 ```
@@ -248,7 +248,7 @@ Each instance becomes an EVAL.yaml with:
 - `workspace.repos[].base_commit` — the commit to reset to before the agent runs
 - `assertions` — `code-grader` tasks that run `FAIL_TO_PASS` and `PASS_TO_PASS` pytest suites inside the container
 
-Run an imported SWE-bench eval against any coding agent target:
+Run an imported SWE-bench eval against any coding agent provider:
 
 ```bash
 # Import one instance
@@ -257,8 +257,8 @@ uv run scripts/import-huggingface.py \
   --limit 1 \
   --output /tmp/swebench-eval/
 
-# Run with a coding agent target
-agentv eval /tmp/swebench-eval/*.EVAL.yaml --target codex-cli
+# Run with a coding agent provider
+agentv eval /tmp/swebench-eval/*.EVAL.yaml --provider codex-cli
 ```
 
 The Docker workspace spins up the pre-built SWE-bench image, checks out `base_commit`, runs the agent to apply a patch, then grades by running the test suite inside the container.
diff --git a/apps/web/src/content/docs/docs/v4.42.4/tools/prepare.mdx b/apps/web/src/content/docs/docs/v4.42.4/tools/prepare.mdx
index 9df429646..099e2efe8 100644
--- a/apps/web/src/content/docs/docs/v4.42.4/tools/prepare.mdx
+++ b/apps/web/src/content/docs/docs/v4.42.4/tools/prepare.mdx
@@ -13,7 +13,7 @@ pagefind: false
 This is the manual-attempt workflow:
 
 ```bash
-agentv prepare evals/foo.eval.yaml --test-id case-1 --target codex-cli --out /tmp/agentv-case-1
+agentv prepare evals/foo.eval.yaml --test-id case-1 --provider codex-cli --out /tmp/agentv-case-1
 ```
 
 The prepared directory contains:
@@ -73,9 +73,9 @@ Use `--response` when the final answer text should be graded independently of th
 
 ## Observability Boundary
 
-`prepare` is not a replacement for live observability. Configure live tracing in the harness or target itself:
+`prepare` is not a replacement for live observability. Configure live tracing in the harness or provider itself:
 
-- Use provider-native settings, target hooks, or environment variables to enable session logs.
+- Use provider-native settings, provider hooks, or environment variables to enable session logs.
 - Emit OpenTelemetry/OpenInference spans from the system under test or provider when you need external trace inspection.
 - For Opik, Phoenix, Langfuse, or another backend, treat their traces as external artifacts that can be imported or projected back into AgentV later.
 
diff --git a/examples/README.md b/examples/README.md
index cc0bb8a3b..53f96627d 100644
--- a/examples/README.md
+++ b/examples/README.md
@@ -92,7 +92,7 @@ example-name/
 │   └── *.md                  # LLM grader prompts (optional)
 ├── scripts/                  # Helper scripts (optional)
 ├── .agentv/
-│   └── targets.yaml          # Target configuration (optional)
+│   └── providers.yaml          # Target configuration (optional)
 ├── package.json              # Dependencies (if using @agentv/sdk)
 └── README.md                 # Example documentation
 ```
diff --git a/examples/contract/.agentv/targets.yaml b/examples/contract/.agentv/providers.yaml
similarity index 76%
rename from examples/contract/.agentv/targets.yaml
rename to examples/contract/.agentv/providers.yaml
index 69305e1db..0314484e0 100644
--- a/examples/contract/.agentv/targets.yaml
+++ b/examples/contract/.agentv/providers.yaml
@@ -1,9 +1,9 @@
-targets:
-  - id: github-models-contract
-    provider: openai
+providers:
+  - id: openai
     api_format: chat
     base_url: https://models.github.ai/inference
     api_key: "{{ env.GH_MODELS_TOKEN }}"
     model: "{{ env.CONTRACT_EVAL_MODEL }}"
     temperature: 0
     max_output_tokens: 512
+    label: github-models-contract
diff --git a/examples/features/agent-skills-evals/README.md b/examples/features/agent-skills-evals/README.md
index 6635c5782..a3d9b2a41 100644
--- a/examples/features/agent-skills-evals/README.md
+++ b/examples/features/agent-skills-evals/README.md
@@ -11,7 +11,7 @@ agent-skills-evals/
 ├── csv-analyzer.EVAL.yaml           # Universal EVAL.yaml format
 ├── csv-analyzer.evals.json          # Transpiled output from EVAL.yaml
 ├── .agentv/
-│   └── targets.yaml                 # Echo provider for no-live-provider testing
+│   └── providers.yaml                 # Echo provider for no-live-provider testing
 └── evals/
     └── files/
         └── sales.csv                # Test fixture
@@ -88,7 +88,7 @@ This produces `csv-analyzer.evals.json` — a skill-creator-compatible file that
 agentv eval csv-analyzer.EVAL.yaml
 
 # Run against a real target
-agentv eval csv-analyzer.EVAL.yaml --target default
+agentv eval csv-analyzer.EVAL.yaml --provider default
 ```
 
 ## Skill setup
@@ -101,7 +101,7 @@ The `csv-analyzer` skill is included in this example under `.claude/skills/csv-a
 
 ```bash
 bun apps/cli/src/cli.ts eval multi-provider-skill-use.EVAL.yaml \
-  --target copilot-cli --targets ../.agentv/targets.yaml
+  --provider copilot-cli --providers ../.agentv/providers.yaml
 ```
 
 The `skill-used` and `not-skill-used` graders automatically handle each provider's tool-call format:
diff --git a/examples/features/autoresearch/README.md b/examples/features/autoresearch/README.md
index 85952a752..533598a51 100644
--- a/examples/features/autoresearch/README.md
+++ b/examples/features/autoresearch/README.md
@@ -12,7 +12,7 @@ Demonstrates the autoresearch optimization loop with a practical scenario.
 This example works like any other eval:
 
 ```bash
-agentv eval EVAL.yaml --experiment autoresearch-classifier --target azure
+agentv eval EVAL.yaml --experiment autoresearch-classifier --provider azure
 ```
 
 To run autoresearch, use the `agentv-bench` skill:
diff --git a/examples/features/basic/README.md b/examples/features/basic/README.md
index de5fc0ddd..7fdc35601 100644
--- a/examples/features/basic/README.md
+++ b/examples/features/basic/README.md
@@ -17,7 +17,7 @@ Demonstrates core AgentV schema features with minimal setup.
 bun agentv eval examples/features/basic/evals/suite.yaml
 
 # With specific target
-bun agentv eval examples/features/basic/evals/suite.yaml --target default
+bun agentv eval examples/features/basic/evals/suite.yaml --provider default
 ```
 
 ## Key Files
diff --git a/examples/features/batch-cli/.agentv/providers.yaml b/examples/features/batch-cli/.agentv/providers.yaml
new file mode 100644
index 000000000..d299122e6
--- /dev/null
+++ b/examples/features/batch-cli/.agentv/providers.yaml
@@ -0,0 +1,11 @@
+providers:
+  - id: cli
+    batch_requests: true
+    verbose: true
+    command: bun run ./scripts/batch-cli-runner.ts --eval-input ./evals/suite.yaml
+      --csv-input ./AmlScreeningInput.csv --output {OUTPUT_FILE}
+    cwd: ..
+    healthcheck:
+      command: bun run ./scripts/batch-cli-runner.ts --healthcheck
+      cwd: ..
+    label: batch_cli
diff --git a/examples/features/batch-cli/.agentv/targets.yaml b/examples/features/batch-cli/.agentv/targets.yaml
deleted file mode 100644
index e6ee7aa53..000000000
--- a/examples/features/batch-cli/.agentv/targets.yaml
+++ /dev/null
@@ -1,13 +0,0 @@
-targets:
-  - id: batch_cli
-    provider: cli
-    batch_requests: true
-    verbose: true
-    # Runs once for the whole panel.
-    # The runner reads ./AmlScreeningInput.csv and writes JSONL to {OUTPUT_FILE}.
-    # NOTE: Do not add quotes around {OUTPUT_FILE}; it is already shell-escaped.
-    command: bun run ./scripts/batch-cli-runner.ts --eval-input ./evals/suite.yaml --csv-input ./AmlScreeningInput.csv --output {OUTPUT_FILE}
-    cwd: ..
-    healthcheck:
-      command: bun run ./scripts/batch-cli-runner.ts --healthcheck
-      cwd: ..
diff --git a/examples/features/batch-cli/README.md b/examples/features/batch-cli/README.md
index 78e4fb267..4721c170d 100644
--- a/examples/features/batch-cli/README.md
+++ b/examples/features/batch-cli/README.md
@@ -52,7 +52,7 @@ Trajectory assertions extract tool calls directly from `output[].tool_calls[]`.
 - `batch-cli-demo.yaml` — Ground truth: tests with inputs and expected outputs
 - `scripts/build-csv-from-eval.ts` — Utilities to convert YAML tests to CSV format (imported by batch-cli-runner.ts)
 - `scripts/batch-cli-runner.ts` — Main batch runner: converts inputs to CSV, processes them, writes actual responses as JSONL
-- `.agentv/targets.yaml` — Defines the `batch_cli` CLI target with request batching enabled
+- `.agentv/providers.yaml` — Defines the `batch_cli` CLI target with request batching enabled
 
 ## Run
 
@@ -63,5 +63,5 @@ cd examples/features/batch-cli
 
 # Run AgentV against the batch CLI target
 # NOTE: This requires the CLI provider to support batching + JSONL batch output.
-bun agentv eval ./evals/suite.yaml --target batch_cli
+bun agentv eval ./evals/suite.yaml --provider batch_cli
 ```
diff --git a/examples/features/benchmark-tooling/README.md b/examples/features/benchmark-tooling/README.md
index 098b21b4f..996d5318f 100644
--- a/examples/features/benchmark-tooling/README.md
+++ b/examples/features/benchmark-tooling/README.md
@@ -130,8 +130,8 @@ Use `agentv results compare` on completed run manifests for pairwise analysis, o
 
 ```bash
 # 1. Run the same eval once per target
-bun agentv eval my-eval.yaml --target gpt-4.1 --experiment model-benchmark
-bun agentv eval my-eval.yaml --target claude-sonnet-4 --experiment model-benchmark
+bun agentv eval my-eval.yaml --provider gpt-4.1 --experiment model-benchmark
+bun agentv eval my-eval.yaml --provider claude-sonnet-4 --experiment model-benchmark
 
 # 2. Compare two completed runs
 bun agentv results compare \
diff --git a/examples/features/copilot-transcript-replay/.agentv/providers.yaml b/examples/features/copilot-transcript-replay/.agentv/providers.yaml
new file mode 100644
index 000000000..6f3983096
--- /dev/null
+++ b/examples/features/copilot-transcript-replay/.agentv/providers.yaml
@@ -0,0 +1,5 @@
+providers:
+  - id: replay
+    transcripts: ../fixtures/copilot-transcript.jsonl
+    source_target: copilot-cli
+    label: copilot-transcript-replay
diff --git a/examples/features/copilot-transcript-replay/.agentv/targets.yaml b/examples/features/copilot-transcript-replay/.agentv/targets.yaml
deleted file mode 100644
index f79ff89b1..000000000
--- a/examples/features/copilot-transcript-replay/.agentv/targets.yaml
+++ /dev/null
@@ -1,8 +0,0 @@
-targets:
-  # Recorded Copilot trajectory replay. Import Copilot CLI events with
-  # `agentv import copilot`, or use the checked-in normalized transcript
-  # fixture below. Graders consume AgentV transcript rows, not raw events.jsonl.
-  - id: copilot-transcript-replay
-    provider: replay
-    transcripts: ../fixtures/copilot-transcript.jsonl
-    source_target: copilot-cli
diff --git a/examples/features/copilot-transcript-replay/README.md b/examples/features/copilot-transcript-replay/README.md
index d832dd607..41622a2ec 100644
--- a/examples/features/copilot-transcript-replay/README.md
+++ b/examples/features/copilot-transcript-replay/README.md
@@ -33,7 +33,7 @@ Import the session into a normalized AgentV transcript:
 agentv import copilot \
   --session-id <uuid> \
   --test-id should-not-trigger-csv-analyzer \
-  --target copilot-cli \
+  --provider copilot-cli \
   -o fixtures/copilot-transcript.jsonl
 ```
 
@@ -43,7 +43,7 @@ example runs.
 ### 2. Run the replay eval
 
 ```bash
-agentv eval evals/skill-use.EVAL.yaml --target copilot-transcript-replay
+agentv eval evals/skill-use.EVAL.yaml --provider copilot-transcript-replay
 ```
 
 The `before_all` hook syncs the agentv-dev plugin skills into the workspace.
diff --git a/examples/features/deterministic-graders/README.md b/examples/features/deterministic-graders/README.md
index fab733c8f..da52ca10f 100644
--- a/examples/features/deterministic-graders/README.md
+++ b/examples/features/deterministic-graders/README.md
@@ -47,7 +47,7 @@ bun run build
 
 ```bash
 # From examples/features
-bun agentv eval deterministic-graders/evals/suite.yaml --target <your-target>
+bun agentv eval deterministic-graders/evals/suite.yaml --provider <your-target>
 ```
 
 ## Standalone Test
diff --git a/examples/features/docker-workspace/.agentv/providers.yaml b/examples/features/docker-workspace/.agentv/providers.yaml
new file mode 100644
index 000000000..0e442eb40
--- /dev/null
+++ b/examples/features/docker-workspace/.agentv/providers.yaml
@@ -0,0 +1,3 @@
+providers:
+  - id: mock
+    label: mock_agent
diff --git a/examples/features/docker-workspace/.agentv/targets.yaml b/examples/features/docker-workspace/.agentv/targets.yaml
deleted file mode 100644
index 7a41cae75..000000000
--- a/examples/features/docker-workspace/.agentv/targets.yaml
+++ /dev/null
@@ -1,3 +0,0 @@
-targets:
-  - id: mock_agent
-    provider: mock
diff --git a/examples/features/document-extraction/.agentv/providers.yaml b/examples/features/document-extraction/.agentv/providers.yaml
new file mode 100644
index 000000000..906990eda
--- /dev/null
+++ b/examples/features/document-extraction/.agentv/providers.yaml
@@ -0,0 +1,8 @@
+providers:
+  - id: cli
+    provider_batching: false
+    verbose: true
+    command: bun run ../mock_extractor.ts {FILES} {OUTPUT_FILE}
+    healthcheck:
+      command: bun --version
+    label: mock_extractor
diff --git a/examples/features/document-extraction/.agentv/targets.yaml b/examples/features/document-extraction/.agentv/targets.yaml
deleted file mode 100644
index a2b1aad9c..000000000
--- a/examples/features/document-extraction/.agentv/targets.yaml
+++ /dev/null
@@ -1,14 +0,0 @@
-targets:
-  - id: mock_extractor
-    provider: cli
-    provider_batching: false
-    verbose: true
-    
-    # Runs the mock invoice extractor for each test case individually
-    # {FILES} is replaced with the input file paths from the test case
-    # {OUTPUT_FILE} is the temporary file path where output should be written
-    command: bun run ../mock_extractor.ts {FILES} {OUTPUT_FILE}
-
-
-    healthcheck:
-      command: bun --version
diff --git a/examples/features/execution-metrics/.agentv/targets.yaml b/examples/features/execution-metrics/.agentv/providers.yaml
similarity index 81%
rename from examples/features/execution-metrics/.agentv/targets.yaml
rename to examples/features/execution-metrics/.agentv/providers.yaml
index ff5d7bee1..742c86178 100644
--- a/examples/features/execution-metrics/.agentv/targets.yaml
+++ b/examples/features/execution-metrics/.agentv/providers.yaml
@@ -1,15 +1,15 @@
-targets:
-  - id: azure-llm
-    provider: azure
+providers:
+  - id: azure
     endpoint: "{{ env.AZURE_OPENAI_ENDPOINT }}"
     api_key: "{{ env.AZURE_OPENAI_API_KEY }}"
     model: "{{ env.AZURE_DEPLOYMENT_NAME }}"
     version: "{{ env.AZURE_OPENAI_API_VERSION }}"
-  - id: mock_metrics_agent
-    provider: cli
+    label: azure-llm
+  - id: cli
     grader_target: azure-llm
     command: bun run ./mock-metrics-agent.ts --prompt {PROMPT} --output {OUTPUT_FILE}
     cwd: ..
     healthcheck:
       command: bun run ./mock-metrics-agent.ts --healthcheck
       cwd: ..
+    label: mock_metrics_agent
diff --git a/examples/features/execution-metrics/README.md b/examples/features/execution-metrics/README.md
index 6ddddc966..9be099553 100644
--- a/examples/features/execution-metrics/README.md
+++ b/examples/features/execution-metrics/README.md
@@ -15,7 +15,7 @@ Demonstrates execution metrics tracking (tokens, cost, latency) in evaluations.
 ```bash
 # From repository root
 cd examples/features
-bun agentv eval execution-metrics/evals/suite.yaml --target mock_metrics_agent
+bun agentv eval execution-metrics/evals/suite.yaml --provider mock_metrics_agent
 ```
 
 ## Setup
diff --git a/examples/features/file-changes-graders/.agentv/providers.yaml b/examples/features/file-changes-graders/.agentv/providers.yaml
new file mode 100644
index 000000000..075bb7a18
--- /dev/null
+++ b/examples/features/file-changes-graders/.agentv/providers.yaml
@@ -0,0 +1,11 @@
+providers:
+  - id: cli
+    command: "bash -c ' printf \"export function add(a: number, b: number): number
+      {\\n  return a + b;\\n}\\n\\nexport function subtract(a: number, b:
+      number): number {\\n  return a - b;\\n}\\n\" > src/calculator.ts && echo
+      \"Added subtract function to calculator.ts\" > {OUTPUT_FILE} '"
+    grader_target: grader
+    label: mock_agent
+  - id: copilot-cli
+    model: claude-haiku-4.5
+    label: copilot_grader
diff --git a/examples/features/file-changes-graders/.agentv/targets.yaml b/examples/features/file-changes-graders/.agentv/targets.yaml
deleted file mode 100644
index 38f7a970a..000000000
--- a/examples/features/file-changes-graders/.agentv/targets.yaml
+++ /dev/null
@@ -1,16 +0,0 @@
-targets:
-  # Mock agent that adds a subtract function to calculator.ts.
-  # Each test gets a fresh copy of workspace-template/ as its CWD.
-  - id: mock_agent
-    provider: cli
-    command: >-
-      bash -c '
-      printf "export function add(a: number, b: number): number {\n  return a + b;\n}\n\nexport function subtract(a: number, b: number): number {\n  return a - b;\n}\n" > src/calculator.ts &&
-      echo "Added subtract function to calculator.ts" > {OUTPUT_FILE}
-      '
-    grader_target: grader
-
-  # Copilot CLI — used as delegated llm-grader target
-  - id: copilot_grader
-    provider: copilot-cli
-    model: claude-haiku-4.5
diff --git a/examples/features/file-changes-with-repos/.agentv/providers.yaml b/examples/features/file-changes-with-repos/.agentv/providers.yaml
new file mode 100644
index 000000000..e9d12569f
--- /dev/null
+++ b/examples/features/file-changes-with-repos/.agentv/providers.yaml
@@ -0,0 +1,7 @@
+providers:
+  - id: cli
+    command: "bash -c ' echo \"Analysis complete: 3 tests passed, 0 failed.\" >
+      report.txt && printf \"\\nexport function add(a: number, b: number):
+      number {\\n  return a + b;\\n}\\n\" >> my-lib/utils.ts && echo \"Done\" >
+      {OUTPUT_FILE} '"
+    label: mock_agent
diff --git a/examples/features/file-changes-with-repos/.agentv/targets.yaml b/examples/features/file-changes-with-repos/.agentv/targets.yaml
deleted file mode 100644
index 55874dcb4..000000000
--- a/examples/features/file-changes-with-repos/.agentv/targets.yaml
+++ /dev/null
@@ -1,11 +0,0 @@
-targets:
-  # Mock agent that writes to the workspace root AND edits a file inside a nested git repo.
-  # Simulates an agent that produces an artifact alongside making code changes.
-  - id: mock_agent
-    provider: cli
-    command: >-
-      bash -c '
-      echo "Analysis complete: 3 tests passed, 0 failed." > report.txt &&
-      printf "\nexport function add(a: number, b: number): number {\n  return a + b;\n}\n" >> my-lib/utils.ts &&
-      echo "Done" > {OUTPUT_FILE}
-      '
diff --git a/examples/features/file-changes/.agentv/providers.yaml b/examples/features/file-changes/.agentv/providers.yaml
new file mode 100644
index 000000000..4857609be
--- /dev/null
+++ b/examples/features/file-changes/.agentv/providers.yaml
@@ -0,0 +1,10 @@
+providers:
+  - id: cli
+    command: "bash -c ' printf \"Hello, AgentV!\\nUpdated by the agent.\\n\" >
+      hello.txt && printf \"{ \\\"version\\\": 2, \\\"debug\\\": true }\\n\" >
+      config.json && mkdir -p src tests && printf \"export const isEmpty = (s:
+      string) => s.length === 0;\\n\" > src/utils.ts && printf \"import { greet
+      } from \\\"../src/main\\\";\\nconsole.log(greet(\\\"World\\\"));\\n\" >
+      tests/main.test.ts && rm -f obsolete.log && echo \"Done: edited 2 files,
+      created 2 files, deleted 1 file.\" > {OUTPUT_FILE} '"
+    label: mock_agent
diff --git a/examples/features/file-changes/.agentv/targets.yaml b/examples/features/file-changes/.agentv/targets.yaml
deleted file mode 100644
index 917521fe8..000000000
--- a/examples/features/file-changes/.agentv/targets.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-targets:
-  # Mock agent that edits, creates, and deletes files in the workspace.
-  # Each test gets a fresh copy of workspace-template/ as its CWD.
-  - id: mock_agent
-    provider: cli
-    command: >-
-      bash -c '
-      printf "Hello, AgentV!\nUpdated by the agent.\n" > hello.txt &&
-      printf "{ \"version\": 2, \"debug\": true }\n" > config.json &&
-      mkdir -p src tests &&
-      printf "export const isEmpty = (s: string) => s.length === 0;\n" > src/utils.ts &&
-      printf "import { greet } from \"../src/main\";\nconsole.log(greet(\"World\"));\n" > tests/main.test.ts &&
-      rm -f obsolete.log &&
-      echo "Done: edited 2 files, created 2 files, deleted 1 file." > {OUTPUT_FILE}
-      '
diff --git a/examples/features/file-transforms/.agentv/targets.yaml b/examples/features/file-transforms/.agentv/providers.yaml
similarity index 77%
rename from examples/features/file-transforms/.agentv/targets.yaml
rename to examples/features/file-transforms/.agentv/providers.yaml
index cd9c4ef2b..fa0eae941 100644
--- a/examples/features/file-transforms/.agentv/targets.yaml
+++ b/examples/features/file-transforms/.agentv/providers.yaml
@@ -1,30 +1,27 @@
 $schema: agentv-targets-v2.2
-targets:
-  - id: file_output
-    provider: cli
+providers:
+  - id: cli
     runtime: host
     config:
       command: bun run .agentv/providers/file-output.ts {OUTPUT_FILE}
       cwd: ..
     grader_target: grader_check
-
-  - id: grader_check
-    provider: cli
+    label: file_output
+  - id: cli
     runtime: host
     config:
       command: bun run .agentv/providers/grader-check.ts {PROMPT_FILE} {OUTPUT_FILE}
       cwd: ..
-
-  - id: file_output_live
-    provider: cli
+    label: grader_check
+  - id: cli
     runtime: host
     config:
       command: bun run .agentv/providers/file-output.ts {OUTPUT_FILE}
       cwd: ..
     grader_target: local-openai-grader
-
-  - id: local-openai-grader
-    provider: openai
+    label: file_output_live
+  - id: openai
     base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
     api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
     model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
+    label: local-openai-grader
diff --git a/examples/features/file-transforms/README.md b/examples/features/file-transforms/README.md
index f1495d644..88611bf0d 100644
--- a/examples/features/file-transforms/README.md
+++ b/examples/features/file-transforms/README.md
@@ -14,7 +14,7 @@ Demonstrates how `default_test.options.transform` turns `ContentFile` outputs in
 
 ```bash
 # From repository root
-bun apps/cli/src/cli.ts eval examples/features/file-transforms/evals/suite.yaml --target file_output
+bun apps/cli/src/cli.ts eval examples/features/file-transforms/evals/suite.yaml --provider file_output
 ```
 
 Expected result: the eval passes because the grader sees the transformed spreadsheet text from `generated/report.xlsx`.
@@ -25,13 +25,13 @@ To run the same file-output path with a live OpenAI-compatible grader:
 LOCAL_OPENAI_PROXY_BASE_URL=http://127.0.0.1:10531/v1 \
 LOCAL_OPENAI_PROXY_API_KEY=dummy-local-key \
 LOCAL_OPENAI_PROXY_MODEL=gpt-5.4-mini \
-bun apps/cli/src/cli.ts eval examples/features/file-transforms/evals/suite.yaml --target file_output_live
+bun apps/cli/src/cli.ts eval examples/features/file-transforms/evals/suite.yaml --provider file_output_live
 ```
 
 ## Key Files
 
 - `evals/suite.yaml` - eval with `default_test.options.transform`
-- `.agentv/targets.yaml` - custom file-producing target and custom grader target
+- `.agentv/providers.yaml` - custom file-producing target and custom grader target
 - `.agentv/providers/file-output.ts` - emits a relative `ContentFile` path
 - `.agentv/providers/grader-check.ts` - passes only when transformed text reaches the grader prompt
 - `scripts/transforms/xlsx-to-csv.ts` - example spreadsheet conversion script
diff --git a/examples/features/functional-grading/.agentv/providers.yaml b/examples/features/functional-grading/.agentv/providers.yaml
new file mode 100644
index 000000000..dabcdb4cc
--- /dev/null
+++ b/examples/features/functional-grading/.agentv/providers.yaml
@@ -0,0 +1,10 @@
+providers:
+  - id: cli
+    command: "bash -c ' printf \"export function add(a: number, b: number): number
+      {\\n  return a + b;\\n}\\n\\nexport function multiply(a: number, b:
+      number): number {\\n  return a * b;\\n}\\n\\nexport function fibonacci(n:
+      number): number {\\n  if (n <= 1) return n;\\n  let a = 0, b = 1;\\n  for
+      (let i = 2; i <= n; i++) {\\n    const tmp = a + b;\\n    a = b;\\n    b =
+      tmp;\\n  }\\n  return b;\\n}\\n\" > src/index.ts && echo \"Implemented
+      add, multiply, and fibonacci functions.\" > {OUTPUT_FILE} '"
+    label: mock_agent
diff --git a/examples/features/functional-grading/.agentv/targets.yaml b/examples/features/functional-grading/.agentv/targets.yaml
deleted file mode 100644
index 9e0f6c1c2..000000000
--- a/examples/features/functional-grading/.agentv/targets.yaml
+++ /dev/null
@@ -1,10 +0,0 @@
-targets:
-  # Mock agent that implements the stub functions in src/index.ts.
-  # Each test gets a fresh copy of workspace-template/ as its CWD.
-  - id: mock_agent
-    provider: cli
-    command: >-
-      bash -c '
-      printf "export function add(a: number, b: number): number {\n  return a + b;\n}\n\nexport function multiply(a: number, b: number): number {\n  return a * b;\n}\n\nexport function fibonacci(n: number): number {\n  if (n <= 1) return n;\n  let a = 0, b = 1;\n  for (let i = 2; i <= n; i++) {\n    const tmp = a + b;\n    a = b;\n    b = tmp;\n  }\n  return b;\n}\n" > src/index.ts &&
-      echo "Implemented add, multiply, and fibonacci functions." > {OUTPUT_FILE}
-      '
diff --git a/examples/features/latency-assertions/.agentv/targets.yaml b/examples/features/latency-assertions/.agentv/providers.yaml
similarity index 80%
rename from examples/features/latency-assertions/.agentv/targets.yaml
rename to examples/features/latency-assertions/.agentv/providers.yaml
index a952d6f43..af1ebcda1 100644
--- a/examples/features/latency-assertions/.agentv/targets.yaml
+++ b/examples/features/latency-assertions/.agentv/providers.yaml
@@ -1,9 +1,9 @@
-targets:
-  - id: mock_latency_agent
-    provider: cli
+providers:
+  - id: cli
     grader_target: grader
     command: bun run ./mock-latency-agent.ts --prompt {PROMPT} --output {OUTPUT_FILE}
     cwd: ..
     healthcheck:
       command: bun run ./mock-latency-agent.ts --healthcheck
       cwd: ..
+    label: mock_latency_agent
diff --git a/examples/features/latency-assertions/README.md b/examples/features/latency-assertions/README.md
index bdcc45390..456fe8868 100644
--- a/examples/features/latency-assertions/README.md
+++ b/examples/features/latency-assertions/README.md
@@ -57,7 +57,7 @@ For trajectory assertions to work, providers must include tool calls:
 npx agentv validate examples/features/latency-assertions/evals/suite.yaml
 
 # With the included mock provider or a real provider that returns tool calls
-npx agentv eval examples/features/latency-assertions/evals/suite.yaml --target mock_latency_agent
+npx agentv eval examples/features/latency-assertions/evals/suite.yaml --provider mock_latency_agent
 ```
 
 ## Best Practices
diff --git a/examples/features/local-cli/.agentv/targets.yaml b/examples/features/local-cli/.agentv/providers.yaml
similarity index 83%
rename from examples/features/local-cli/.agentv/targets.yaml
rename to examples/features/local-cli/.agentv/providers.yaml
index 48daa619b..b3a57439f 100644
--- a/examples/features/local-cli/.agentv/targets.yaml
+++ b/examples/features/local-cli/.agentv/providers.yaml
@@ -1,6 +1,5 @@
-targets:
-  - id: local_cli
-    provider: cli
+providers:
+  - id: cli
     grader_target: grader
     command: uv run ./mock_cli.py --prompt {PROMPT} {FILES} --output {OUTPUT_FILE}
     files_format: --file {path}
@@ -8,3 +7,4 @@ targets:
     healthcheck:
       command: uv run ./mock_cli.py --healthcheck
       cwd: ..
+    label: local_cli
diff --git a/examples/features/local-cli/README.md b/examples/features/local-cli/README.md
index e166c1ce8..2dfdc31b0 100644
--- a/examples/features/local-cli/README.md
+++ b/examples/features/local-cli/README.md
@@ -19,4 +19,4 @@ bun agentv eval examples/features/local-cli/evals/suite.yaml
 ## Key Files
 
 - `evals/suite.yaml` - Test cases with file attachments
-- `.agentv/targets.yaml` - Local CLI target configuration
+- `.agentv/providers.yaml` - Local CLI target configuration
diff --git a/examples/features/readme-quickstart/.agentv/providers.yaml b/examples/features/readme-quickstart/.agentv/providers.yaml
new file mode 100644
index 000000000..bc1fa2e62
--- /dev/null
+++ b/examples/features/readme-quickstart/.agentv/providers.yaml
@@ -0,0 +1,17 @@
+providers:
+  - id: openai
+    runtime: host
+    config:
+      api_format: chat
+      base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
+      api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
+      model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
+    label: local-openai
+  - id: openai
+    runtime: host
+    config:
+      api_format: chat
+      base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
+      api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
+      model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
+    label: local-openai-grader
diff --git a/examples/features/readme-quickstart/.agentv/targets.yaml b/examples/features/readme-quickstart/.agentv/targets.yaml
deleted file mode 100644
index 09f039a78..000000000
--- a/examples/features/readme-quickstart/.agentv/targets.yaml
+++ /dev/null
@@ -1,17 +0,0 @@
-- id: local-openai
-  provider: openai
-  runtime: host
-  config:
-    api_format: chat
-    base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
-    api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
-    model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
-
-- id: local-openai-grader
-  provider: openai
-  runtime: host
-  config:
-    api_format: chat
-    base_url: "{{ env.LOCAL_OPENAI_PROXY_BASE_URL }}"
-    api_key: "{{ env.LOCAL_OPENAI_PROXY_API_KEY }}"
-    model: "{{ env.LOCAL_OPENAI_PROXY_MODEL }}"
diff --git a/examples/features/readme-quickstart/README.md b/examples/features/readme-quickstart/README.md
index 4ba71673f..6d76dd243 100644
--- a/examples/features/readme-quickstart/README.md
+++ b/examples/features/readme-quickstart/README.md
@@ -6,13 +6,13 @@ It includes a composable `.agentv/config.yaml` that decomposes the base config
 graph into direct field refs:
 
 ```yaml
-targets: file://targets.yaml
+targets: file://providers.yaml
 tests: file://tests.yaml
 defaults: file://defaults.yaml
 ```
 
 Each referenced file contains that field's value directly, such as a bare target
-array in `.agentv/targets.yaml` and a bare defaults object in
+array in `.agentv/providers.yaml` and a bare defaults object in
 `.agentv/defaults.yaml`. A grader is not a separate kind of entity — it is a
 target listed under `targets` like any other, selected for the grading role
 via `defaults.grader`.
diff --git a/examples/features/readme-quickstart/targets.yaml b/examples/features/readme-quickstart/providers.yaml
similarity index 100%
rename from examples/features/readme-quickstart/targets.yaml
rename to examples/features/readme-quickstart/providers.yaml
diff --git a/examples/features/rubric/README.md b/examples/features/rubric/README.md
index b3fbc7f13..8dd65e903 100644
--- a/examples/features/rubric/README.md
+++ b/examples/features/rubric/README.md
@@ -15,7 +15,7 @@ Demonstrates rubric-based evaluation with weights, required flags, and auto-gene
 
 ```bash
 # From repository root
-bun agentv eval examples/features/rubric/evals/suite.yaml --target default
+bun agentv eval examples/features/rubric/evals/suite.yaml --provider default
 ```
 
 ## Key Files
diff --git a/examples/features/scenarios/README.md b/examples/features/scenarios/README.md
index 71da892e6..cf6eb1e74 100644
--- a/examples/features/scenarios/README.md
+++ b/examples/features/scenarios/README.md
@@ -81,13 +81,13 @@ compares it to the reference answer from the config row.
 # From repository root
 bun apps/cli/src/cli.ts validate examples/features/scenarios/evals/suite.yaml
 bun apps/cli/src/cli.ts eval run examples/features/scenarios/evals/suite.yaml \
-  --targets examples/features/scenarios/targets.yaml \
-  --target translation-cli
+  --providers examples/features/scenarios/providers.yaml \
+  --provider translation-cli
 ```
 
 ## Key Files
 
 - `evals/suite.yaml` - Main eval with a file-backed scenario matrix
 - `evals/scenarios/translation-matrix.yaml` - Scenario file loaded by glob
-- `targets.yaml` - Deterministic CLI target for running the example
+- `providers.yaml` - Deterministic CLI target for running the example
 - `scripts/translation-target.mjs` - Prompt-to-translation target script
diff --git a/examples/features/scenarios/providers.yaml b/examples/features/scenarios/providers.yaml
new file mode 100644
index 000000000..ef5a73ab5
--- /dev/null
+++ b/examples/features/scenarios/providers.yaml
@@ -0,0 +1,4 @@
+providers:
+  - id: cli
+    label: translation-cli
+    command: node ../scripts/translation-target.mjs {PROMPT_FILE} {OUTPUT_FILE}
diff --git a/examples/features/scenarios/targets.yaml b/examples/features/scenarios/targets.yaml
deleted file mode 100644
index 68542ef31..000000000
--- a/examples/features/scenarios/targets.yaml
+++ /dev/null
@@ -1,4 +0,0 @@
-providers:
-  - id: cli
-    label: translation-cli
-    command: "node ../scripts/translation-target.mjs {PROMPT_FILE} {OUTPUT_FILE}"
diff --git a/examples/features/script-grader-sdk/.agentv/targets.yaml b/examples/features/script-grader-sdk/.agentv/providers.yaml
similarity index 73%
rename from examples/features/script-grader-sdk/.agentv/targets.yaml
rename to examples/features/script-grader-sdk/.agentv/providers.yaml
index f19f5281f..08cb695ad 100644
--- a/examples/features/script-grader-sdk/.agentv/targets.yaml
+++ b/examples/features/script-grader-sdk/.agentv/providers.yaml
@@ -1,10 +1,11 @@
-targets:
-  - id: local_cli
-    provider: cli
+providers:
+  - id: cli
     grader_target: grader
-    command: uv run ../local-cli/mock_cli.py --prompt {PROMPT} {FILES} --output {OUTPUT_FILE}
+    command: uv run ../local-cli/mock_cli.py --prompt {PROMPT} {FILES} --output
+      {OUTPUT_FILE}
     files_format: --file {path}
     cwd: ..
     healthcheck:
       command: uv run ../local-cli/mock_cli.py --healthcheck
       cwd: ..
+    label: local_cli
diff --git a/examples/features/script-grader-sdk/README.md b/examples/features/script-grader-sdk/README.md
index 5dab6f026..c5256f7bd 100644
--- a/examples/features/script-grader-sdk/README.md
+++ b/examples/features/script-grader-sdk/README.md
@@ -44,10 +44,10 @@ From the repository root:
 
 ```bash
 cd examples/features
-bun agentv eval script-grader-sdk/evals/suite.yaml --target local_cli
+bun agentv eval script-grader-sdk/evals/suite.yaml --provider local_cli
 ```
 
-This requires a CLI target named `local_cli` configured in `.agentv/targets.yaml`.
+This requires a CLI target named `local_cli` configured in `.agentv/providers.yaml`.
 
 ## API
 
diff --git a/examples/features/script-grader-with-llm-calls/README.md b/examples/features/script-grader-with-llm-calls/README.md
index e1e9b1dd6..445c6e51c 100644
--- a/examples/features/script-grader-with-llm-calls/README.md
+++ b/examples/features/script-grader-with-llm-calls/README.md
@@ -231,10 +231,10 @@ The `createTargetClient()` function reads these automatically.
 # From the agentv monorepo root:
 
 # Run contextual precision evaluation
-agentv eval examples/features/script-grader-with-llm-calls/evals/contextual-precision.eval.yaml --target gemini-llm
+agentv eval examples/features/script-grader-with-llm-calls/evals/contextual-precision.eval.yaml --provider gemini-llm
 
 # Run contextual recall evaluation
-agentv eval examples/features/script-grader-with-llm-calls/evals/contextual-recall.eval.yaml --target gemini-llm
+agentv eval examples/features/script-grader-with-llm-calls/evals/contextual-recall.eval.yaml --provider gemini-llm
 ```
 
 ### Expected Results
diff --git a/examples/features/sdk-eval-authoring/.agentv/providers.yaml b/examples/features/sdk-eval-authoring/.agentv/providers.yaml
new file mode 100644
index 000000000..91391abfc
--- /dev/null
+++ b/examples/features/sdk-eval-authoring/.agentv/providers.yaml
@@ -0,0 +1,4 @@
+providers:
+  - id: mock
+    response: Hello from the mock target
+    label: mock-sdk
diff --git a/examples/features/sdk-eval-authoring/.agentv/targets.yaml b/examples/features/sdk-eval-authoring/.agentv/targets.yaml
deleted file mode 100644
index d73ec1087..000000000
--- a/examples/features/sdk-eval-authoring/.agentv/targets.yaml
+++ /dev/null
@@ -1,4 +0,0 @@
-targets:
-  - id: mock-sdk
-    provider: mock
-    response: Hello from the mock target
diff --git a/examples/features/sdk-eval-authoring/README.md b/examples/features/sdk-eval-authoring/README.md
index 36ff3bbeb..6d5bc1512 100644
--- a/examples/features/sdk-eval-authoring/README.md
+++ b/examples/features/sdk-eval-authoring/README.md
@@ -12,7 +12,7 @@ Demonstrates authoring an explicit `*.eval.ts` file with the `EvalConfig` type f
 ## Files
 
 - `evals/greeting.eval.ts` — the TypeScript eval config
-- `.agentv/targets.yaml` — local mock target for a zero-credential run
+- `.agentv/providers.yaml` — local mock target for a zero-credential run
 - `fixtures/per-test-note.md` — attached input file used by the suite
 
 ## How to Run
diff --git a/examples/features/sdk-programmatic-api/evaluate.ts b/examples/features/sdk-programmatic-api/evaluate.ts
index fc24f9b68..a8523a322 100644
--- a/examples/features/sdk-programmatic-api/evaluate.ts
+++ b/examples/features/sdk-programmatic-api/evaluate.ts
@@ -5,7 +5,7 @@
  * The inline config mirrors the canonical YAML surface with TypeScript-friendly names.
  *
  * Run: bun run evaluate.ts
- * (Uses 'default' target from .agentv/targets.yaml and .env credentials)
+ * (Uses 'default' target from .agentv/providers.yaml and .env credentials)
  */
 import { evaluate } from '@agentv/sdk';
 
diff --git a/examples/features/sdk-python/.agentv/providers.yaml b/examples/features/sdk-python/.agentv/providers.yaml
new file mode 100644
index 000000000..6e76f82a0
--- /dev/null
+++ b/examples/features/sdk-python/.agentv/providers.yaml
@@ -0,0 +1,9 @@
+providers:
+  - id: cli
+    command: python3 -c "import argparse; parser = argparse.ArgumentParser();
+      parser.add_argument('--prompt', required=True);
+      parser.add_argument('--output', required=True); args, _ =
+      parser.parse_known_args(); open(args.output, 'w',
+      encoding='utf-8').write(args.prompt)" --prompt {PROMPT} --output
+      {OUTPUT_FILE}
+    label: local_cli
diff --git a/examples/features/sdk-python/.agentv/targets.yaml b/examples/features/sdk-python/.agentv/targets.yaml
deleted file mode 100644
index 028cbb9c0..000000000
--- a/examples/features/sdk-python/.agentv/targets.yaml
+++ /dev/null
@@ -1,4 +0,0 @@
-targets:
-  - id: local_cli
-    provider: cli
-    command: python3 -c "import argparse; parser = argparse.ArgumentParser(); parser.add_argument('--prompt', required=True); parser.add_argument('--output', required=True); args, _ = parser.parse_known_args(); open(args.output, 'w', encoding='utf-8').write(args.prompt)" --prompt {PROMPT} --output {OUTPUT_FILE}
diff --git a/examples/features/sdk-python/README.md b/examples/features/sdk-python/README.md
index 2db20f80f..935236c93 100644
--- a/examples/features/sdk-python/README.md
+++ b/examples/features/sdk-python/README.md
@@ -37,5 +37,5 @@ uv run python scripts/check_expected_output.py < sample-grader-input.json
 To run the generated eval through AgentV from the repository root:
 
 ```bash
-bun apps/cli/src/cli.ts eval examples/features/sdk-python/evals/suite.yaml --target local_cli
+bun apps/cli/src/cli.ts eval examples/features/sdk-python/evals/suite.yaml --provider local_cli
 ```
diff --git a/examples/features/suite-level-input/.agentv/providers.yaml b/examples/features/suite-level-input/.agentv/providers.yaml
new file mode 100644
index 000000000..530d3c7f0
--- /dev/null
+++ b/examples/features/suite-level-input/.agentv/providers.yaml
@@ -0,0 +1,3 @@
+providers:
+  - id: mock
+    label: llm
diff --git a/examples/features/suite-level-input/.agentv/targets.yaml b/examples/features/suite-level-input/.agentv/targets.yaml
deleted file mode 100644
index c8b0847a5..000000000
--- a/examples/features/suite-level-input/.agentv/targets.yaml
+++ /dev/null
@@ -1,3 +0,0 @@
-targets:
-  - id: llm
-    provider: mock
diff --git a/examples/features/tool-evaluation-plugins/.agentv/providers.yaml b/examples/features/tool-evaluation-plugins/.agentv/providers.yaml
new file mode 100644
index 000000000..530d3c7f0
--- /dev/null
+++ b/examples/features/tool-evaluation-plugins/.agentv/providers.yaml
@@ -0,0 +1,3 @@
+providers:
+  - id: mock
+    label: llm
diff --git a/examples/features/tool-evaluation-plugins/.agentv/targets.yaml b/examples/features/tool-evaluation-plugins/.agentv/targets.yaml
deleted file mode 100644
index c8b0847a5..000000000
--- a/examples/features/tool-evaluation-plugins/.agentv/targets.yaml
+++ /dev/null
@@ -1,3 +0,0 @@
-targets:
-  - id: llm
-    provider: mock
diff --git a/examples/features/tool-evaluation-plugins/README.md b/examples/features/tool-evaluation-plugins/README.md
index fa1450317..8413cd2f1 100644
--- a/examples/features/tool-evaluation-plugins/README.md
+++ b/examples/features/tool-evaluation-plugins/README.md
@@ -39,7 +39,7 @@ graders:
 
 ```bash
 cd examples/features/tool-evaluation-plugins
-bun agentv eval evals/suite.yaml --target <your-target>
+bun agentv eval evals/suite.yaml --provider <your-target>
 ```
 
 ## Output
diff --git a/examples/features/trajectory-assertions-advanced/.agentv/targets.yaml b/examples/features/trajectory-assertions-advanced/.agentv/providers.yaml
similarity index 66%
rename from examples/features/trajectory-assertions-advanced/.agentv/targets.yaml
rename to examples/features/trajectory-assertions-advanced/.agentv/providers.yaml
index 3debcb94e..e13225540 100644
--- a/examples/features/trajectory-assertions-advanced/.agentv/targets.yaml
+++ b/examples/features/trajectory-assertions-advanced/.agentv/providers.yaml
@@ -1,9 +1,10 @@
-targets:
-  - id: static_trace
-    provider: cli
+providers:
+  - id: cli
     grader_target: grader
-    command: bun run ./cat-trace.ts --trace ./static-trace.json --prompt {PROMPT} --output {OUTPUT_FILE}
+    command: bun run ./cat-trace.ts --trace ./static-trace.json --prompt {PROMPT}
+      --output {OUTPUT_FILE}
     cwd: ..
     healthcheck:
       command: bun run ./cat-trace.ts --healthcheck
       cwd: ..
+    label: static_trace
diff --git a/examples/features/trajectory-assertions-advanced/README.md b/examples/features/trajectory-assertions-advanced/README.md
index c38dc0a36..c4e3c3f2c 100644
--- a/examples/features/trajectory-assertions-advanced/README.md
+++ b/examples/features/trajectory-assertions-advanced/README.md
@@ -15,7 +15,7 @@ Demonstrates Promptfoo-compatible trajectory assertions combined with expected o
 ```bash
 # From repository root
 cd examples/features
-bun agentv eval trajectory-assertions-advanced/evals/trace-file-demo.eval.yaml --target static_trace
+bun agentv eval trajectory-assertions-advanced/evals/trace-file-demo.eval.yaml --provider static_trace
 ```
 
 ## Setup
diff --git a/examples/features/trajectory-assertions-simple/.agentv/targets.yaml b/examples/features/trajectory-assertions-simple/.agentv/providers.yaml
similarity index 81%
rename from examples/features/trajectory-assertions-simple/.agentv/targets.yaml
rename to examples/features/trajectory-assertions-simple/.agentv/providers.yaml
index ca4672f0d..63739e19e 100644
--- a/examples/features/trajectory-assertions-simple/.agentv/targets.yaml
+++ b/examples/features/trajectory-assertions-simple/.agentv/providers.yaml
@@ -1,9 +1,9 @@
-targets:
-  - id: mock_agent
-    provider: cli
+providers:
+  - id: cli
     grader_target: grader
     command: bun run ./mock-agent.ts --prompt {PROMPT} --output {OUTPUT_FILE}
     cwd: ..
     healthcheck:
       command: bun run ./mock-agent.ts --healthcheck
       cwd: ..
+    label: mock_agent
diff --git a/examples/features/trajectory-assertions-simple/README.md b/examples/features/trajectory-assertions-simple/README.md
index ff503355f..54ddd8809 100644
--- a/examples/features/trajectory-assertions-simple/README.md
+++ b/examples/features/trajectory-assertions-simple/README.md
@@ -15,7 +15,7 @@ Demonstrates Promptfoo-compatible trajectory assertions with different matching
 ```bash
 # From repository root
 cd examples/features
-bun agentv eval trajectory-assertions-simple/evals/suite.yaml --target mock_agent
+bun agentv eval trajectory-assertions-simple/evals/suite.yaml --provider mock_agent
 ```
 
 ## Setup
diff --git a/examples/features/trend/README.md b/examples/features/trend/README.md
index 20c915706..8b2b993f1 100644
--- a/examples/features/trend/README.md
+++ b/examples/features/trend/README.md
@@ -36,7 +36,7 @@ cp -R sample-runs/* .agentv/results/default/
 Then run:
 
 ```bash
-bun ../../../apps/cli/src/cli.ts trend --last 3 --suite code-review --target claude-sonnet
+bun ../../../apps/cli/src/cli.ts trend --last 3 --suite code-review --provider claude-sonnet
 ```
 
 Expected output:
@@ -74,7 +74,7 @@ bun ../../../apps/cli/src/cli.ts trend \
   sample-runs/2026-03-08T10-00-00-000Z \
   sample-runs/2026-03-15T10-00-00-000Z \
   --suite code-review \
-  --target claude-sonnet
+  --provider claude-sonnet
 ```
 
 ## CI Gate Example
@@ -87,7 +87,7 @@ bun ../../../apps/cli/src/cli.ts trend \
   sample-runs/2026-03-08T10-00-00-000Z \
   sample-runs/2026-03-15T10-00-00-000Z \
   --suite code-review \
-  --target claude-sonnet \
+  --provider claude-sonnet \
   --fail-on-degrading \
   --slope-threshold 0.01
 ```
diff --git a/examples/features/trial-output-consistency/README.md b/examples/features/trial-output-consistency/README.md
index 6a99e89f0..27917d41f 100644
--- a/examples/features/trial-output-consistency/README.md
+++ b/examples/features/trial-output-consistency/README.md
@@ -65,7 +65,7 @@ assert:
 bun agentv validate examples/features/trial-output-consistency/evals/suite.yaml
 
 # Run a specific test with a configured target
-bun agentv eval examples/features/trial-output-consistency/evals/suite.yaml --test-id high-consistency --target <your-target>
+bun agentv eval examples/features/trial-output-consistency/evals/suite.yaml --test-id high-consistency --provider <your-target>
 ```
 
 ## Extending
diff --git a/examples/features/vitest-workspace-grader/.agentv/targets.yaml b/examples/features/vitest-workspace-grader/.agentv/providers.yaml
similarity index 90%
rename from examples/features/vitest-workspace-grader/.agentv/targets.yaml
rename to examples/features/vitest-workspace-grader/.agentv/providers.yaml
index 8f9d22c43..f69adc186 100644
--- a/examples/features/vitest-workspace-grader/.agentv/targets.yaml
+++ b/examples/features/vitest-workspace-grader/.agentv/providers.yaml
@@ -1,6 +1,5 @@
-targets:
-  - id: mock_agent
-    provider: cli
+providers:
+  - id: cli
     command: |
       bash -c '
       mkdir -p app &&
@@ -18,3 +17,4 @@ targets:
       EOF
       echo "Updated the welcome banner." > {OUTPUT_FILE}
       '
+    label: mock_agent
diff --git a/examples/features/vitest-workspace-grader/README.md b/examples/features/vitest-workspace-grader/README.md
index 3bd600b9d..b0d3cfff0 100644
--- a/examples/features/vitest-workspace-grader/README.md
+++ b/examples/features/vitest-workspace-grader/README.md
@@ -6,7 +6,7 @@ Demonstrates the preferred deterministic workspace grader path: write normal Vit
 
 - `graders/welcome-banner.test.ts`: plain Vitest verifier that reads `app/page.tsx`
 - `evals/suite.yaml`: eval case that runs the verifier through `agentv eval <verifier.test.ts>`
-- `.agentv/targets.yaml`: mock CLI target that updates the workspace
+- `.agentv/providers.yaml`: mock CLI target that updates the workspace
 
 ## Run
 
@@ -15,7 +15,7 @@ From this example directory:
 ```bash
 bun install
 cd ../../..
-bun apps/cli/src/cli.ts eval examples/features/vitest-workspace-grader/evals/suite.yaml --target mock_agent
+bun apps/cli/src/cli.ts eval examples/features/vitest-workspace-grader/evals/suite.yaml --provider mock_agent
 ```
 
 ## Pattern
diff --git a/examples/features/workspace-artifact/.agentv/providers.yaml b/examples/features/workspace-artifact/.agentv/providers.yaml
new file mode 100644
index 000000000..8b573ee9c
--- /dev/null
+++ b/examples/features/workspace-artifact/.agentv/providers.yaml
@@ -0,0 +1,7 @@
+providers:
+  - id: cli
+    command: bash -c ' mkdir -p outputs && printf
+      "metric,value,status\ncoverage,87.3,pass\nlatency_p99_ms,142,pass\nerror_rate,0.02,pass\n"
+      > outputs/report.csv && echo "Generated outputs/report.csv" >
+      {OUTPUT_FILE} '
+    label: mock_csv_agent
diff --git a/examples/features/workspace-artifact/.agentv/targets.yaml b/examples/features/workspace-artifact/.agentv/targets.yaml
deleted file mode 100644
index ca8b6097d..000000000
--- a/examples/features/workspace-artifact/.agentv/targets.yaml
+++ /dev/null
@@ -1,11 +0,0 @@
-targets:
-  # Mock CLI agent that writes a CSV report to outputs/report.csv under workspace_path.
-  # Simulates what a real agent (e.g. Copilot) would do when asked to generate a report.
-  - id: mock_csv_agent
-    provider: cli
-    command: >-
-      bash -c '
-      mkdir -p outputs &&
-      printf "metric,value,status\ncoverage,87.3,pass\nlatency_p99_ms,142,pass\nerror_rate,0.02,pass\n" > outputs/report.csv &&
-      echo "Generated outputs/report.csv" > {OUTPUT_FILE}
-      '
diff --git a/examples/red-team/README.md b/examples/red-team/README.md
index 094afd6e3..2ab313a1e 100644
--- a/examples/red-team/README.md
+++ b/examples/red-team/README.md
@@ -42,7 +42,7 @@ examples/red-team/
 
 ```bash
 # Single suite against your own target
-agentv eval examples/red-team/suites/llm01-prompt-injection.yaml --target your-target
+agentv eval examples/red-team/suites/llm01-prompt-injection.yaml --provider your-target
 
 # Filter by governance tag (any built-in jq pipeline works because the metadata
 # is present verbatim in the JSONL):
diff --git a/examples/red-team/archetypes/coding-agent/README.md b/examples/red-team/archetypes/coding-agent/README.md
index 68f142887..7e2aac2b3 100644
--- a/examples/red-team/archetypes/coding-agent/README.md
+++ b/examples/red-team/archetypes/coding-agent/README.md
@@ -94,11 +94,11 @@ v5.4** technique pages (<https://atlas.mitre.org/>, public) for AML.T0051
 ```bash
 # A single suite against your own coding-agent target
 agentv eval examples/red-team/archetypes/coding-agent/suites/destructive-git.eval.yaml \
-  --target your-coder-target
+  --provider your-coder-target
 
 # All suites at once
 agentv eval examples/red-team/archetypes/coding-agent/suites/ \
-  --target your-coder-target
+  --provider your-coder-target
 ```
 
 The pack is opt-in. Copy what you need.
diff --git a/examples/red-team/archetypes/customer-facing-agent/README.md b/examples/red-team/archetypes/customer-facing-agent/README.md
index 806b81ff4..085b1ecac 100644
--- a/examples/red-team/archetypes/customer-facing-agent/README.md
+++ b/examples/red-team/archetypes/customer-facing-agent/README.md
@@ -85,11 +85,11 @@ CC-BY-SA 4.0) for the IDs we tag against.
 
 ```bash
 agentv eval examples/red-team/archetypes/customer-facing-agent/suites/pii-cross-customer-disclosure.eval.yaml \
-  --target your-support-target
+  --provider your-support-target
 
 # All suites at once
 agentv eval examples/red-team/archetypes/customer-facing-agent/suites/ \
-  --target your-support-target
+  --provider your-support-target
 ```
 
 The pack is opt-in. Copy what you need.
diff --git a/examples/showcase/bug-fix-benchmark/README.md b/examples/showcase/bug-fix-benchmark/README.md
index 85f4de754..8320400d1 100644
--- a/examples/showcase/bug-fix-benchmark/README.md
+++ b/examples/showcase/bug-fix-benchmark/README.md
@@ -40,7 +40,7 @@ Compares four configurations on identical bug fix tasks:
 agentv eval evals/bug-fixes.eval.yaml --workers 3
 
 # Override the target from the CLI
-agentv eval evals/bug-fixes.eval.yaml --target claude-baseline --workers 2
+agentv eval evals/bug-fixes.eval.yaml --provider claude-baseline --workers 2
 ```
 
 ### 2. Compare results
@@ -62,7 +62,7 @@ The eval includes a real bug from the AgentV repo:
 
 ## How Variants Work
 
-The `claude` target from the repo root `.agentv/targets.yaml` is used as the
+The `claude` target from the repo root `.agentv/providers.yaml` is used as the
 base. The eval file uses an eval-local target object with hooks:
 
 ```yaml
diff --git a/examples/showcase/cross-repo-sync/.agentv/providers.yaml b/examples/showcase/cross-repo-sync/.agentv/providers.yaml
new file mode 100644
index 000000000..e20180269
--- /dev/null
+++ b/examples/showcase/cross-repo-sync/.agentv/providers.yaml
@@ -0,0 +1,9 @@
+providers:
+  - id: cli
+    command: bash mock-agent.sh {PROMPT} {OUTPUT_FILE}
+    label: mock_agent
+  - id: claude
+    label: claude_agent
+  - id: copilot-cli
+    model: claude-haiku-4.5
+    label: copilot_agent
diff --git a/examples/showcase/cross-repo-sync/.agentv/targets.yaml b/examples/showcase/cross-repo-sync/.agentv/targets.yaml
deleted file mode 100644
index 6b573588e..000000000
--- a/examples/showcase/cross-repo-sync/.agentv/targets.yaml
+++ /dev/null
@@ -1,11 +0,0 @@
-targets:
-  - id: mock_agent
-    provider: cli
-    command: bash mock-agent.sh {PROMPT} {OUTPUT_FILE}
-
-  - id: claude_agent
-    provider: claude
-
-  - id: copilot_agent
-    provider: copilot-cli
-    model: claude-haiku-4.5
diff --git a/examples/showcase/cross-repo-sync/README.md b/examples/showcase/cross-repo-sync/README.md
index d7c664d98..f121173e0 100644
--- a/examples/showcase/cross-repo-sync/README.md
+++ b/examples/showcase/cross-repo-sync/README.md
@@ -44,6 +44,6 @@ bun agentv eval ./evals/suite.yaml
 │   ├── reset.ts                   # after_each: git reset
 │   └── validate-sync.ts           # script grader
 ├── .agentv/
-│   └── targets.yaml               # Mock CLI agent
+│   └── providers.yaml               # Mock CLI agent
 └── package.json
 ```
diff --git a/examples/showcase/export-screening/.agentv/providers.yaml b/examples/showcase/export-screening/.agentv/providers.yaml
new file mode 100644
index 000000000..3ce7863aa
--- /dev/null
+++ b/examples/showcase/export-screening/.agentv/providers.yaml
@@ -0,0 +1,4 @@
+providers:
+  - id: mock
+    response: '{"riskLevel": "High", "reasoning": "Mock response for testing"}'
+    label: default
diff --git a/examples/showcase/export-screening/.agentv/targets.yaml b/examples/showcase/export-screening/.agentv/targets.yaml
deleted file mode 100644
index cd54298e2..000000000
--- a/examples/showcase/export-screening/.agentv/targets.yaml
+++ /dev/null
@@ -1,10 +0,0 @@
-# Export Screening Mock Targets
-# This file overrides the parent targets.yaml for local testing without LLM calls.
-# Customize the 'response' field to test different scenarios.
-
-targets:
-  # Mock target for testing eval flow without LLM calls
-  # Default: returns High risk - customize to test other scenarios
-  - id: default
-    provider: mock
-    response: '{"riskLevel": "High", "reasoning": "Mock response for testing"}'
diff --git a/examples/showcase/export-screening/README.md b/examples/showcase/export-screening/README.md
index 0087b4296..a6daf614d 100644
--- a/examples/showcase/export-screening/README.md
+++ b/examples/showcase/export-screening/README.md
@@ -30,7 +30,7 @@ export-screening/
 │   ├── validate_risk_output.ts         # JSON validator + accuracy checker
 │   └── ci_check.ts                     # CI/CD threshold checker
 └── .agentv/
-    └── targets.yaml                    # (optional) target configuration
+    └── providers.yaml                    # (optional) target configuration
 ```
 
 ## Running the Evaluation
diff --git a/examples/showcase/multi-model-benchmark/README.md b/examples/showcase/multi-model-benchmark/README.md
index 2466c619e..a6feb11da 100644
--- a/examples/showcase/multi-model-benchmark/README.md
+++ b/examples/showcase/multi-model-benchmark/README.md
@@ -26,7 +26,7 @@ multi-model-benchmark/
 
 ## Prerequisites
 
-1. Configure targets in `.agentv/targets.yaml` at the repository root. The example commands use `copilot`, `claude`, and `gemini-llm` — these must be defined with valid provider credentials.
+1. Configure providers in `.agentv/providers.yaml` at the repository root. The example commands use `copilot`, `claude`, and `gemini-llm` — these must be defined with valid provider credentials.
 2. Install dependencies: `bun install`
 
 ## Running the Evaluation
@@ -37,11 +37,11 @@ From the repository root:
 # Run once per target. Use the same experiment label so Dashboard analytics
 # can group the completed runs.
 bun agentv eval examples/showcase/multi-model-benchmark/evals/benchmark.eval.yaml \
-  --target copilot --experiment multi-model-benchmark
+  --provider copilot --experiment multi-model-benchmark
 bun agentv eval examples/showcase/multi-model-benchmark/evals/benchmark.eval.yaml \
-  --target claude --experiment multi-model-benchmark
+  --provider claude --experiment multi-model-benchmark
 bun agentv eval examples/showcase/multi-model-benchmark/evals/benchmark.eval.yaml \
-  --target gemini-llm --experiment multi-model-benchmark
+  --provider gemini-llm --experiment multi-model-benchmark
 ```
 
 ### Cost & Safety
@@ -53,7 +53,7 @@ To run against a single target first:
 ```bash
 # Test with one model before running the other targets
 bun agentv eval examples/showcase/multi-model-benchmark/evals/benchmark.eval.yaml \
-  --target copilot
+  --provider copilot
 ```
 
 ## Comparing Models
@@ -104,12 +104,12 @@ Pairwise Summary:
 ### 1. Target Selection
 
 The eval file names one default target. Run the same eval with different
-`--target` values to compare models:
+`--provider` values to compare models:
 
 ```bash
-agentv eval evals/benchmark.eval.yaml --target copilot
-agentv eval evals/benchmark.eval.yaml --target claude
-agentv eval evals/benchmark.eval.yaml --target gemini-llm
+agentv eval evals/benchmark.eval.yaml --provider copilot
+agentv eval evals/benchmark.eval.yaml --provider claude
+agentv eval evals/benchmark.eval.yaml --provider gemini-llm
 ```
 
 ### 2. Weighted Graders
@@ -183,11 +183,11 @@ benchmark.eval.yaml
 
 ### Adding a model
 
-Add a new target to `.agentv/targets.yaml`, then run the same eval with `--target <name>` and the same `--experiment` label.
+Add a new target to `.agentv/providers.yaml`, then run the same eval with `--provider <name>` and the same `--experiment` label.
 
 ```bash
 bun agentv eval examples/showcase/multi-model-benchmark/evals/benchmark.eval.yaml \
-  --target my_new_model --experiment multi-model-benchmark
+  --provider my_new_model --experiment multi-model-benchmark
 ```
 
 ### Adding a grader
diff --git a/examples/showcase/offline-grader-benchmark/.agentv/targets.yaml b/examples/showcase/offline-grader-benchmark/.agentv/providers.yaml
similarity index 69%
rename from examples/showcase/offline-grader-benchmark/.agentv/targets.yaml
rename to examples/showcase/offline-grader-benchmark/.agentv/providers.yaml
index 93ca8172e..abad5e339 100644
--- a/examples/showcase/offline-grader-benchmark/.agentv/targets.yaml
+++ b/examples/showcase/offline-grader-benchmark/.agentv/providers.yaml
@@ -1,9 +1,10 @@
-targets:
-  - id: fixture_replay
-    provider: cli
-    command: bun run ./scripts/replay-fixture-output.ts --prompt {PROMPT} --output {OUTPUT_FILE}
+providers:
+  - id: cli
+    command: bun run ./scripts/replay-fixture-output.ts --prompt {PROMPT} --output
+      {OUTPUT_FILE}
     cwd: ..
     grader_target: grader
     healthcheck:
       command: bun run ./scripts/replay-fixture-output.ts --healthcheck
       cwd: ..
+    label: fixture_replay
diff --git a/examples/showcase/offline-grader-benchmark/README.md b/examples/showcase/offline-grader-benchmark/README.md
index 197a64222..ee0ab072e 100644
--- a/examples/showcase/offline-grader-benchmark/README.md
+++ b/examples/showcase/offline-grader-benchmark/README.md
@@ -13,7 +13,7 @@ It uses existing AgentV primitives:
 
 ```text
 offline-grader-benchmark/
-├── .agentv/targets.yaml                  # Replay target + three illustrative low-cost grader targets
+├── .agentv/providers.yaml                  # Replay target + three illustrative low-cost grader targets
 ├── README.md
 ├── evals/
 │   ├── setup-a.eval.yaml                 # Grader setup A
@@ -59,7 +59,7 @@ The example ships with three illustrative low-cost graders:
 - `grader_claude_haiku` via OpenRouter model `anthropic/claude-haiku-4.5`
 - `grader_gemini_flash` via OpenRouter model `google/gemini-3-flash-preview`
 
-Edit `.agentv/targets.yaml` if your local environment uses different deployment names or model IDs.
+Edit `.agentv/providers.yaml` if your local environment uses different deployment names or model IDs.
 
 ## No-API-key smoke test
 
diff --git a/examples/showcase/trace-evaluation/.agentv/targets.yaml b/examples/showcase/trace-evaluation/.agentv/providers.yaml
similarity index 59%
rename from examples/showcase/trace-evaluation/.agentv/targets.yaml
rename to examples/showcase/trace-evaluation/.agentv/providers.yaml
index 4c2a1511a..d099e32f3 100644
--- a/examples/showcase/trace-evaluation/.agentv/targets.yaml
+++ b/examples/showcase/trace-evaluation/.agentv/providers.yaml
@@ -1,16 +1,14 @@
-targets:
-  - id: live_coding_agent
-    provider: codex-cli
+providers:
+  - id: codex-cli
     model: gpt-5
     timeout_seconds: 300
-
-  - id: replay_coding_agent
-    provider: replay
+    label: live_coding_agent
+  - id: replay
     fixtures: ../fixtures/replay-target-output.jsonl
     suite: trace-evaluation-showcase
     source_target: live_coding_agent
-
-  - id: replay_imported_codex_transcript
-    provider: replay
+    label: replay_coding_agent
+  - id: replay
     transcripts: ../fixtures/imported-codex-transcript.jsonl
     source_target: codex
+    label: replay_imported_codex_transcript
diff --git a/examples/showcase/trace-evaluation/README.md b/examples/showcase/trace-evaluation/README.md
index cbf958997..0b0830a1e 100644
--- a/examples/showcase/trace-evaluation/README.md
+++ b/examples/showcase/trace-evaluation/README.md
@@ -11,7 +11,7 @@ This showcase demonstrates trace evaluation as target substitution:
 
 ```text
 trace-evaluation/
-├── .agentv/targets.yaml
+├── .agentv/providers.yaml
 ├── evals/
 │   ├── coding-agent-replay.eval.yaml
 │   └── transcript-import.eval.yaml
@@ -33,7 +33,7 @@ From the repository root:
 ```bash
 bun apps/cli/src/cli.ts eval \
   examples/showcase/trace-evaluation/evals/coding-agent-replay.eval.yaml \
-  --target replay_coding_agent \
+  --provider replay_coding_agent \
   --output /tmp/agentv-trace-showcase-replay-run
 ```
 
@@ -81,7 +81,7 @@ target and `--record-replay`:
 ```bash
 bun apps/cli/src/cli.ts eval \
   examples/showcase/trace-evaluation/evals/coding-agent-replay.eval.yaml \
-  --target live_coding_agent \
+  --provider live_coding_agent \
   --record-replay examples/showcase/trace-evaluation/fixtures/replay-target-output.jsonl \
   --output /tmp/agentv-trace-showcase-live-run
 ```
diff --git a/examples/showcase/trace-evaluation/scripts/prove-replay.ts b/examples/showcase/trace-evaluation/scripts/prove-replay.ts
index cf212922c..72e1b1719 100644
--- a/examples/showcase/trace-evaluation/scripts/prove-replay.ts
+++ b/examples/showcase/trace-evaluation/scripts/prove-replay.ts
@@ -17,7 +17,7 @@ function runEval(): void {
       'apps/cli/src/cli.ts',
       'eval',
       'examples/showcase/trace-evaluation/evals/coding-agent-replay.eval.yaml',
-      '--target',
+      '--provider',
       'replay_coding_agent',
       '--output',
       resultDir,
diff --git a/packages/core/src/evaluation/evaluate.ts b/packages/core/src/evaluation/evaluate.ts
index 17b9ad318..a3541437f 100644
--- a/packages/core/src/evaluation/evaluate.ts
+++ b/packages/core/src/evaluation/evaluate.ts
@@ -743,7 +743,7 @@ function computeSummary(
   };
 }
 
-const TARGET_FILE_CANDIDATES = ['.agentv/providers.yaml', '.agentv/providers.yml'] as const;
+const PROVIDER_FILE_CANDIDATES = ['.agentv/providers.yaml', '.agentv/providers.yml'] as const;
 
 /**
  * Auto-discover the 'default' provider from providers.yaml in the repo tree.
@@ -753,7 +753,7 @@ async function discoverDefaultTarget(repoRoot: string): Promise<TargetDefinition
   const chain = buildDirectoryChain(path.join(cwd, '_placeholder'), repoRoot);
 
   for (const dir of chain) {
-    for (const candidate of TARGET_FILE_CANDIDATES) {
+    for (const candidate of PROVIDER_FILE_CANDIDATES) {
       const targetsPath = path.join(dir, candidate);
       if (!existsSync(targetsPath)) continue;
       try {