@ai-sdk-tool/harness@1.3.2

Fixes

• env.ts: moved .env file discovery side-effects out of module top level into loadDotEnvFilesIfAvailable() helper. The module now loads safely in edge runtimes (Cloudflare Workers, Vercel Edge) without node:fs access. Callers that relied on the automatic .env loading must explicitly call loadDotEnvFilesIfAvailable() from their Node.js entry point.

plugsuits@2.3.5

Patch Changes

• e937dc7: Keep env-file loading compatible with Node 18, preserve URL validation for shared AI endpoints, and align CEA's default context limit with the shared AI configuration.

• 5bb3997: Update direct and transitive dependency resolutions across the monorepo, including AI SDK packages, tooling, TypeScript, and runtime adapters. Raise the declared Node.js support floor to 22.19.0 to match upgraded runtime dependencies such as undici 8.

• 8b1919c: Persist user-level agent preferences across sessions (e.g. /translate, /reasoning-mode, /tool-fallback in CEA; /reasoning in minimal-agent) so toggles set in the TUI survive process restarts.

  • Harness: new generic PreferencesStore<T> abstraction with FilePreferencesStore (single atomic JSON document), InMemoryPreferencesStore, LayeredPreferencesStore (configurable merge + write layer), and shallowMergePreferences helper. Exposed from the package root and a new @ai-sdk-tool/harness/preferences subpath. Intentionally separate from SnapshotStore because preferences are app/user-scoped while snapshots are session-scoped.
  • Harness: new one-line helper createLayeredPreferences({ appName, validate }) that returns { store, userStore, workspaceStore, patch, paths } backed by ~/.${appName}/settings.json (user layer) and ./.${appName}/settings.json (workspace layer, write target). patch(partial) handles the load-merge-save flow on the workspace layer so consumers don't have to reimplement it. Fully customizable paths, merge strategy, and validator — but the common case is now a single call.
  • Harness: new createTogglePreferenceCommand and createEnumPreferenceCommand factories that collapse the typical "parse args → validate → mutate runtime → persist" slash-command boilerplate into a declarative config object. Typed field: keyof T ensures persistence goes to the right preference key. Supports aliases, custom truthy/falsy words, custom parser, custom validator, and custom enabled/disabled messages.
  • CEA: createUserPreferencesStore() now delegates to createLayeredPreferences and also exposes the full harness bundle alongside the existing public fields. /translate is migrated to createTogglePreferenceCommand; the CEA-local createToggleCommand factory is deleted as dead code. /translate now awaits persistence (previously fire-and-forget), so the command response confirms the disk write.
  • CEA: /reasoning-mode and /tool-fallback continue to use the shared preferences-persistence singleton for now (they have domain-specific selectable-modes logic that is not a good fit for the generic factory yet). configurePreferencesPersistence now accepts an optional bundle argument so future migrations can reuse the harness factories.
  • CEA startup: persisted preferences are applied to AgentManager before CLI flags. Explicit CLI flags (--no-translate, --reasoning-mode on, --tool-fallback, --toolcall-mode) still win for the current process but no longer overwrite the persisted file — they are one-shot overrides only. resolveSharedConfig now accepts rawArgs so callers can distinguish explicit flags from defaults.
  • minimal-agent: gains a /reasoning <on|off> slash command that toggles providerOptions.openai.reasoningEffort via the onBeforeTurn hook and persists the value through createLayeredPreferences. The command definition is a single 10-line factory call thanks to createTogglePreferenceCommand. Header subtitle shows the live reasoning state. Preferences are stored at ~/.minimal-agent/settings.json and ./.minimal-agent/settings.json (separate from CEA so the two agents' defaults don't collide).

• 8b1919c: Harden the user-preferences persistence layer against failure and concurrency, discovered during manual QA of PR #119 and flagged by the Oracle reviewer.

  • Harness: FilePreferencesStore.clear() now deletes the file with rmSync instead of writing "{}". Previously clear() left an empty {} on disk, and un-validated stores would return {} from load() instead of null, breaking the contract shared with InMemoryPreferencesStore. Consumers using a validator were unaffected in practice because empty objects already validated to null.
  • Harness: createLayeredPreferences().patch() now serializes calls through a promise queue. Previous implementation was a plain read-modify-write, so two concurrent patch() calls could both read the same stale workspace JSON and one update would be silently dropped. Concurrency tests cover both different-field and same-field writes.
  • Harness: createTogglePreferenceCommand and createEnumPreferenceCommand now persist before mutating runtime. If disk persistence fails, the runtime state is never touched and the command returns { success: false, message: "Failed to persist …" }. If the runtime set() throws after a successful persist, the disk write is rolled back to the previous value. This eliminates the class of "disk and runtime disagree" bugs across all factory-backed commands.
  • CEA: patchWorkspacePreferences now queues writes per-store via a WeakMap-keyed promise chain, matching the harness bundle semantics. This protects the shim path used by /reasoning-mode and /tool-fallback from the same lost-update race that previously affected the factory path.
  • CEA: applyPersistedPreferencesToAgentManager and applySharedConfigToAgentManager extracted from main.ts into packages/cea/src/entrypoints/preferences-startup.ts. Now pure functions taking the agent manager and store as arguments, so startup wiring is testable in isolation. Existing main.ts call sites updated to pass dependencies explicitly.
  • CEA: minimal-agent gains vitest devDep, test script, vitest.config.ts, and excludes *.test.ts / vitest.config.ts from the TypeScript build output. citty was a dead devDependency — removed.
  • Tests added:
    • Harness preferences-store.test.ts: clear() ⇒ load() null, clear() removes file on disk, clear() no-op on missing file, cross-instance restart, three concurrent-patch scenarios (different fields, same field, queue recovery after failure).
    • Harness preference-commands.test.ts: persist-first contract — toggle + enum commands do NOT mutate runtime when persistence fails, toggle command rolls back disk when runtime set throws, toggle command returns success only when BOTH disk and runtime succeed.
    • CEA user-preferences.test.ts: concurrent patchWorkspacePreferences preserves all fields, same-field is last-writer-wins, save errors bubble to caller.
    • CEA preferences-persistence.test.ts (new): singleton lifecycle, custom onError handler fires on save failure, concurrent persistPreferencePatch preserves all fields.
    • CEA reasoning-mode.test.ts (new): 7 tests covering status report, valid transitions, invalid mode rejection, persistence, sibling field preservation, no-op "already using", graceful behavior without persistence configured.
    • CEA tool-fallback.test.ts (new): 7 tests in the same shape.
    • CEA preferences-startup.test.ts (new): integration tests for applyPersistedPreferencesToAgentManager and applySharedConfigToAgentManager covering all three fields, CLI-vs-persisted precedence, and the invariant that CLI flags never write to the store.
    • minimal-agent preferences.test.ts (new): 7 tests covering path expectations, schema validation, round-trip, and the /reasoning onBeforeTurn integration (toggle → closure mutation → providerOptions change).
  • Test counts:
    • Harness: 673 → 699 (+26)
    • CEA: 526 → 563 (+37)
    • minimal-agent: 0 → 7 (+7)
    • Total workspace: 1283 → 1353 (+70)

• 496ffdb: Surface the "prompt processing" state that previously looked frozen, and fix follow-up correctness gaps found during post-implementation review.

  • Harness: new LoopHooks.onStreamStart / onFirstStreamPart hooks wrap the agent.stream() call site so consumers driving turns through runAgentLoop can react to the prompt-processing latency gap. onFirstStreamPart receives the current stream part as its first argument (TextStreamPart<ToolSet>) so consumers can inspect part.type to filter framing chunks (start, text-start, …) from visible content. TextStreamPart is re-exported from the harness root for convenience. Docstring clarifies that the TUI has its own independent onStreamStart on AgentTUIConfig.
  • TUI: shows a Processing... loader during turn preparation and transitions to Working... once the LLM request is in flight. The startup token probe is now non-blocking (fire-and-forget) so the editor accepts input immediately; the context-usage footer starts from the estimated count and quietly upgrades to the real value. During a blocking compaction the foreground loader temporarily switches to Compacting... and restores the previous label when the block ends, so users see the actual reason for a long wait. text-start stream parts are now treated as visible, clearing the streaming loader as soon as the assistant view mounts (no more empty-view flicker).
  • Headless: emits a turn-start lifecycle annotation and a matching onStreamStart callback before each LLM request; the event is dropped from trajectory.json (transient UX signal, no step_id) so persisted consumers see identical output. The event fires exactly once per logical turn — overflow and no-output retries no longer re-emit it. New tests cover normal ordering, new-turn vs intermediate-step phases, retry single-emission, and non-persistence in trajectory.json.
  • Headless: the persisted schema_version is corrected from the internal ATIF-v1.6 label to the actual current Harbor spec version ATIF-v1.4 (<https://www.harborframework.com/docs/agents/traject...

@ai-sdk-tool/tui@3.1.1

Patch Changes

• 5bb3997: Update direct and transitive dependency resolutions across the monorepo, including AI SDK packages, tooling, TypeScript, and runtime adapters. Raise the declared Node.js support floor to 22.19.0 to match upgraded runtime dependencies such as undici 8.

• 496ffdb: Surface the "prompt processing" state that previously looked frozen, and fix follow-up correctness gaps found during post-implementation review.

  • Harness: new LoopHooks.onStreamStart / onFirstStreamPart hooks wrap the agent.stream() call site so consumers driving turns through runAgentLoop can react to the prompt-processing latency gap. onFirstStreamPart receives the current stream part as its first argument (TextStreamPart<ToolSet>) so consumers can inspect part.type to filter framing chunks (start, text-start, …) from visible content. TextStreamPart is re-exported from the harness root for convenience. Docstring clarifies that the TUI has its own independent onStreamStart on AgentTUIConfig.
  • TUI: shows a Processing... loader during turn preparation and transitions to Working... once the LLM request is in flight. The startup token probe is now non-blocking (fire-and-forget) so the editor accepts input immediately; the context-usage footer starts from the estimated count and quietly upgrades to the real value. During a blocking compaction the foreground loader temporarily switches to Compacting... and restores the previous label when the block ends, so users see the actual reason for a long wait. text-start stream parts are now treated as visible, clearing the streaming loader as soon as the assistant view mounts (no more empty-view flicker).
  • Headless: emits a turn-start lifecycle annotation and a matching onStreamStart callback before each LLM request; the event is dropped from trajectory.json (transient UX signal, no step_id) so persisted consumers see identical output. The event fires exactly once per logical turn — overflow and no-output retries no longer re-emit it. New tests cover normal ordering, new-turn vs intermediate-step phases, retry single-emission, and non-persistence in trajectory.json.
  • Headless: the persisted schema_version is corrected from the internal ATIF-v1.6 label to the actual current Harbor spec version ATIF-v1.4 (https://www.harborframework.com/docs/agents/trajectory-format). Documentation across packages/headless/AGENTS.md, packages/headless/README.md, and packages/cea/benchmark/AGENTS.md now separates the internal JSONL streaming protocol (which carries lifecycle annotations such as approval, compaction, interrupt, turn-start) from the ATIF-v1.4 trajectory that TrajectoryCollector writes to disk.
  • Headless: StepMetrics gains the remaining ATIF-v1.4 optional fields (logprobs, prompt_token_ids, completion_token_ids) and TrajectoryJson.final_metrics now aggregates total_cost_usd. TrajectoryJson.extra is typed as a closed record of exactly the three ATIF persistence buckets (approval_events, compaction_events, interrupt_events); new lifecycle types must extend the interface explicitly so the Harbor persistence contract stays type-enforced.
  • CEA: the --atif CLI help text and the benchmark pipeline now reference ATIF-v1.4 (matching the corrected schema_version). The bundled packages/cea/benchmark/test_trajectory.py validator now calls Harbor's official TrajectoryValidator when harbor is importable and falls back to a stricter local shape check otherwise; it enforces per-step metric shapes and rejects bool values where ATIF requires a real number.
  • Addressed PR review feedback:
    • turn-start and onStreamStart now fire strictly after agent.stream() successfully returns, so stream-creation failures no longer produce a false "stream started" signal (reported by Gemini, Codex, and Cubic reviewers).
    • The background startup usage probe is serialized against per-turn probes by a generation token; a stale startup probe can no longer overwrite newer usage data and skew context-pressure metrics.
    • The blocking-compaction spinner swap only stashes the original foreground label on first entry and only restores it when the foreground loader is still live, eliminating both the "Compacting..." wording sticking after unblock and the "Processing..." spinner resurrecting after the first stream part arrived.
    • Restored the post-onSetup updateHeader() call that was accidentally dropped when the startup probe became non-blocking, so any header/footer state that onSetup initialises renders immediately instead of waiting for the first probe to resolve.
    • The bundled Python ATIF validator (test_trajectory.py) no longer accepts bool values where ATIF v1.4 requires a real number — isinstance(True, int) is True in Python, so the old check let invalid metric payloads slip through. Added _is_real_number / _is_real_int helpers that exclude bool.
    • Observer hooks (onStreamStart, onFirstStreamPart) no longer abort a valid stream when the callback throws. Errors are logged via console.error and swallowed in the harness loop, headless runner, and TUI session loop, with the contract documented on LoopHooks.
    • Repaired a regression where LoopHooks.onToolCall had silently dropped out of the public LoopHooks interface while still being destructured inside runAgentLoop. The field is restored to its original signature; consumers that already relied on it are unaffected, and the destructuring now type-checks again.
    • Corrected the LoopHooks.onFirstStreamPart signature as a pre-adoption fix (Cubic P2): the previous (context) => void shape promised in its docstring that consumers could filter on part type, but the callback never received the part. The signature now passes (part: TextStreamPart<ToolSet>, context) so consumers can actually inspect part.type. Zero existing consumers were found across the monorepo (the hook was introduced earlier in this PR), so this is a type-only correction with no runtime migration. New regression tests in loop.test.ts cover single-fire semantics, per-iteration firing, empty-stream skip, and observer-error isolation.
  • Pinned the ATIF v1.4 compliance contract in-source: trajectory-collector.ts, TrajectoryJson, AtifStep, TrajectoryEvent, collectTrajectoryEvent, and runHeadless now carry module/interface-level JSDoc spelling out the Harbor spec version, the allowed steps[*].source values, the extra.* persistence rule, and the stream-vs-snapshot boundary. packages/headless/AGENTS.md gains an "ATIF v1.4 COMPLIANCE" section listing the same invariants, and the atif-events.test.ts suite now declares itself as the executable compliance contract. These are docs-only, but they turn future spec drifts into obvious code-review red flags instead of silent regressions.
  • Review cycle 1 follow-ups (Oracle + Gemini + Codex + Cubic + CodeRabbit):
    • Guarded TrajectoryCollector.writeTo against persisting an invalid zero-step trajectory (Harbor's own validator rejects steps: []). The method now returns boolean — true when a file was written, false when the write was intentionally skipped to keep trajectory.json ATIF-v1.4 compliant.
    • Moved the TUI showLoader("Processing...") call inside the stream-turn try/finally so a thrown prepareMessages (or onBeforeTurn/usage probe/compaction check) no longer leaves the spinner stuck on screen.
    • Tightened the startup usage-probe guard: in addition to the generation token, measureUsageIfAvailable now captures messageHistory.getRevision() at call time and drops its result when the history has mutated mid-probe, preventing stale empty-message usage from overwriting per-turn measurements.
    • Narrowed TrajectoryJson.extra to the three canonical lifecycle buckets (approval_events, compaction_events, interrupt_events) by dropping the Record<string, unknown> intersection. New lifecycle types must now extend the interface explicitly, keeping the ATIF persistence contract type-enforced.
    • Hardened the Python validator: _is_real_number now rejects NaN, Infinity, and -Infinity (all of which json.loads will happily produce from non-strict JSON) via an explicit math.isfinite check.
    • Corrected documentation drift across packages/headless/AGENTS.md, packages/headless/README.md, packages/headless/src/types.ts, packages/headless/src/trajectory-collector.ts, and the root AGENTS.md: approval/compaction/interrupt are persisted under trajectory.extra.*, not JSONL-only; only turn-start and error are transient.
    • Regression test added for the writeTo zero-step guard: does not write an invalid zero-step trajectory when the stream fails before any step.
  • Review cycle 2 follow-ups (Oracle re-audit):
    • Headless measureUsageIfAvailable now carries the same generation + revision guards the TUI already had. A slow background probe that resolves after a compaction or a newer per-turn probe no longer overwrites fresh usage data.
    • ATIF v1.4 step source contract aligned across code, Python validator, and benchmark docs: user, agent, and system are all permitted (Harbor v1.2+). Previous divergence between AtifStep.source and test_trajectory.py's valid_sources = {user, agent} is resolved.
    • Root README.md headless event list now includes turn-start and points at Harbor's ATIF-v1.4 schema for the persisted trajectory.

• f523de9: Bump outdated dependencies to their latest releases: @ai-sdk-tool/parser 4.1.21, vitest 4.1.5, and @mariozechner/pi-tui 0.68.1. Align the @ai-sdk-tool/tui peer range for @mariozechner/pi-tui to ^0.68.1 and update createAliasAwareAutocompleteProvider to the new async autocomp...

@ai-sdk-tool/headless@3.1.1

Patch Changes

• 5bb3997: Update direct and transitive dependency resolutions across the monorepo, including AI SDK packages, tooling, TypeScript, and runtime adapters. Raise the declared Node.js support floor to 22.19.0 to match upgraded runtime dependencies such as undici 8.

• 496ffdb: Surface the "prompt processing" state that previously looked frozen, and fix follow-up correctness gaps found during post-implementation review.

  • Harness: new LoopHooks.onStreamStart / onFirstStreamPart hooks wrap the agent.stream() call site so consumers driving turns through runAgentLoop can react to the prompt-processing latency gap. onFirstStreamPart receives the current stream part as its first argument (TextStreamPart<ToolSet>) so consumers can inspect part.type to filter framing chunks (start, text-start, …) from visible content. TextStreamPart is re-exported from the harness root for convenience. Docstring clarifies that the TUI has its own independent onStreamStart on AgentTUIConfig.
  • TUI: shows a Processing... loader during turn preparation and transitions to Working... once the LLM request is in flight. The startup token probe is now non-blocking (fire-and-forget) so the editor accepts input immediately; the context-usage footer starts from the estimated count and quietly upgrades to the real value. During a blocking compaction the foreground loader temporarily switches to Compacting... and restores the previous label when the block ends, so users see the actual reason for a long wait. text-start stream parts are now treated as visible, clearing the streaming loader as soon as the assistant view mounts (no more empty-view flicker).
  • Headless: emits a turn-start lifecycle annotation and a matching onStreamStart callback before each LLM request; the event is dropped from trajectory.json (transient UX signal, no step_id) so persisted consumers see identical output. The event fires exactly once per logical turn — overflow and no-output retries no longer re-emit it. New tests cover normal ordering, new-turn vs intermediate-step phases, retry single-emission, and non-persistence in trajectory.json.
  • Headless: the persisted schema_version is corrected from the internal ATIF-v1.6 label to the actual current Harbor spec version ATIF-v1.4 (https://www.harborframework.com/docs/agents/trajectory-format). Documentation across packages/headless/AGENTS.md, packages/headless/README.md, and packages/cea/benchmark/AGENTS.md now separates the internal JSONL streaming protocol (which carries lifecycle annotations such as approval, compaction, interrupt, turn-start) from the ATIF-v1.4 trajectory that TrajectoryCollector writes to disk.
  • Headless: StepMetrics gains the remaining ATIF-v1.4 optional fields (logprobs, prompt_token_ids, completion_token_ids) and TrajectoryJson.final_metrics now aggregates total_cost_usd. TrajectoryJson.extra is typed as a closed record of exactly the three ATIF persistence buckets (approval_events, compaction_events, interrupt_events); new lifecycle types must extend the interface explicitly so the Harbor persistence contract stays type-enforced.
  • CEA: the --atif CLI help text and the benchmark pipeline now reference ATIF-v1.4 (matching the corrected schema_version). The bundled packages/cea/benchmark/test_trajectory.py validator now calls Harbor's official TrajectoryValidator when harbor is importable and falls back to a stricter local shape check otherwise; it enforces per-step metric shapes and rejects bool values where ATIF requires a real number.
  • Addressed PR review feedback:
    • turn-start and onStreamStart now fire strictly after agent.stream() successfully returns, so stream-creation failures no longer produce a false "stream started" signal (reported by Gemini, Codex, and Cubic reviewers).
    • The background startup usage probe is serialized against per-turn probes by a generation token; a stale startup probe can no longer overwrite newer usage data and skew context-pressure metrics.
    • The blocking-compaction spinner swap only stashes the original foreground label on first entry and only restores it when the foreground loader is still live, eliminating both the "Compacting..." wording sticking after unblock and the "Processing..." spinner resurrecting after the first stream part arrived.
    • Restored the post-onSetup updateHeader() call that was accidentally dropped when the startup probe became non-blocking, so any header/footer state that onSetup initialises renders immediately instead of waiting for the first probe to resolve.
    • The bundled Python ATIF validator (test_trajectory.py) no longer accepts bool values where ATIF v1.4 requires a real number — isinstance(True, int) is True in Python, so the old check let invalid metric payloads slip through. Added _is_real_number / _is_real_int helpers that exclude bool.
    • Observer hooks (onStreamStart, onFirstStreamPart) no longer abort a valid stream when the callback throws. Errors are logged via console.error and swallowed in the harness loop, headless runner, and TUI session loop, with the contract documented on LoopHooks.
    • Repaired a regression where LoopHooks.onToolCall had silently dropped out of the public LoopHooks interface while still being destructured inside runAgentLoop. The field is restored to its original signature; consumers that already relied on it are unaffected, and the destructuring now type-checks again.
    • Corrected the LoopHooks.onFirstStreamPart signature as a pre-adoption fix (Cubic P2): the previous (context) => void shape promised in its docstring that consumers could filter on part type, but the callback never received the part. The signature now passes (part: TextStreamPart<ToolSet>, context) so consumers can actually inspect part.type. Zero existing consumers were found across the monorepo (the hook was introduced earlier in this PR), so this is a type-only correction with no runtime migration. New regression tests in loop.test.ts cover single-fire semantics, per-iteration firing, empty-stream skip, and observer-error isolation.
  • Pinned the ATIF v1.4 compliance contract in-source: trajectory-collector.ts, TrajectoryJson, AtifStep, TrajectoryEvent, collectTrajectoryEvent, and runHeadless now carry module/interface-level JSDoc spelling out the Harbor spec version, the allowed steps[*].source values, the extra.* persistence rule, and the stream-vs-snapshot boundary. packages/headless/AGENTS.md gains an "ATIF v1.4 COMPLIANCE" section listing the same invariants, and the atif-events.test.ts suite now declares itself as the executable compliance contract. These are docs-only, but they turn future spec drifts into obvious code-review red flags instead of silent regressions.
  • Review cycle 1 follow-ups (Oracle + Gemini + Codex + Cubic + CodeRabbit):
    • Guarded TrajectoryCollector.writeTo against persisting an invalid zero-step trajectory (Harbor's own validator rejects steps: []). The method now returns boolean — true when a file was written, false when the write was intentionally skipped to keep trajectory.json ATIF-v1.4 compliant.
    • Moved the TUI showLoader("Processing...") call inside the stream-turn try/finally so a thrown prepareMessages (or onBeforeTurn/usage probe/compaction check) no longer leaves the spinner stuck on screen.
    • Tightened the startup usage-probe guard: in addition to the generation token, measureUsageIfAvailable now captures messageHistory.getRevision() at call time and drops its result when the history has mutated mid-probe, preventing stale empty-message usage from overwriting per-turn measurements.
    • Narrowed TrajectoryJson.extra to the three canonical lifecycle buckets (approval_events, compaction_events, interrupt_events) by dropping the Record<string, unknown> intersection. New lifecycle types must now extend the interface explicitly, keeping the ATIF persistence contract type-enforced.
    • Hardened the Python validator: _is_real_number now rejects NaN, Infinity, and -Infinity (all of which json.loads will happily produce from non-strict JSON) via an explicit math.isfinite check.
    • Corrected documentation drift across packages/headless/AGENTS.md, packages/headless/README.md, packages/headless/src/types.ts, packages/headless/src/trajectory-collector.ts, and the root AGENTS.md: approval/compaction/interrupt are persisted under trajectory.extra.*, not JSONL-only; only turn-start and error are transient.
    • Regression test added for the writeTo zero-step guard: does not write an invalid zero-step trajectory when the stream fails before any step.
  • Review cycle 2 follow-ups (Oracle re-audit):
    • Headless measureUsageIfAvailable now carries the same generation + revision guards the TUI already had. A slow background probe that resolves after a compaction or a newer per-turn probe no longer overwrites fresh usage data.
    • ATIF v1.4 step source contract aligned across code, Python validator, and benchmark docs: user, agent, and system are all permitted (Harbor v1.2+). Previous divergence between AtifStep.source and test_trajectory.py's valid_sources = {user, agent} is resolved.
    • Root README.md headless event list now includes turn-start and points at Harbor's ATIF-v1.4 schema for the persisted trajectory.

@ai-sdk-tool/harness@1.3.1

Patch Changes

• e937dc7: Keep env-file loading compatible with Node 18, preserve URL validation for shared AI endpoints, and align CEA's default context limit with the shared AI configuration.

• 5bb3997: Update direct and transitive dependency resolutions across the monorepo, including AI SDK packages, tooling, TypeScript, and runtime adapters. Raise the declared Node.js support floor to 22.19.0 to match upgraded runtime dependencies such as undici 8.

• 8b1919c: Persist user-level agent preferences across sessions (e.g. /translate, /reasoning-mode, /tool-fallback in CEA; /reasoning in minimal-agent) so toggles set in the TUI survive process restarts.

  • Harness: new generic PreferencesStore<T> abstraction with FilePreferencesStore (single atomic JSON document), InMemoryPreferencesStore, LayeredPreferencesStore (configurable merge + write layer), and shallowMergePreferences helper. Exposed from the package root and a new @ai-sdk-tool/harness/preferences subpath. Intentionally separate from SnapshotStore because preferences are app/user-scoped while snapshots are session-scoped.
  • Harness: new one-line helper createLayeredPreferences({ appName, validate }) that returns { store, userStore, workspaceStore, patch, paths } backed by ~/.${appName}/settings.json (user layer) and ./.${appName}/settings.json (workspace layer, write target). patch(partial) handles the load-merge-save flow on the workspace layer so consumers don't have to reimplement it. Fully customizable paths, merge strategy, and validator — but the common case is now a single call.
  • Harness: new createTogglePreferenceCommand and createEnumPreferenceCommand factories that collapse the typical "parse args → validate → mutate runtime → persist" slash-command boilerplate into a declarative config object. Typed field: keyof T ensures persistence goes to the right preference key. Supports aliases, custom truthy/falsy words, custom parser, custom validator, and custom enabled/disabled messages.
  • CEA: createUserPreferencesStore() now delegates to createLayeredPreferences and also exposes the full harness bundle alongside the existing public fields. /translate is migrated to createTogglePreferenceCommand; the CEA-local createToggleCommand factory is deleted as dead code. /translate now awaits persistence (previously fire-and-forget), so the command response confirms the disk write.
  • CEA: /reasoning-mode and /tool-fallback continue to use the shared preferences-persistence singleton for now (they have domain-specific selectable-modes logic that is not a good fit for the generic factory yet). configurePreferencesPersistence now accepts an optional bundle argument so future migrations can reuse the harness factories.
  • CEA startup: persisted preferences are applied to AgentManager before CLI flags. Explicit CLI flags (--no-translate, --reasoning-mode on, --tool-fallback, --toolcall-mode) still win for the current process but no longer overwrite the persisted file — they are one-shot overrides only. resolveSharedConfig now accepts rawArgs so callers can distinguish explicit flags from defaults.
  • minimal-agent: gains a /reasoning <on|off> slash command that toggles providerOptions.openai.reasoningEffort via the onBeforeTurn hook and persists the value through createLayeredPreferences. The command definition is a single 10-line factory call thanks to createTogglePreferenceCommand. Header subtitle shows the live reasoning state. Preferences are stored at ~/.minimal-agent/settings.json and ./.minimal-agent/settings.json (separate from CEA so the two agents' defaults don't collide).

• 8b1919c: Harden the user-preferences persistence layer against failure and concurrency, discovered during manual QA of PR #119 and flagged by the Oracle reviewer.

  • Harness: FilePreferencesStore.clear() now deletes the file with rmSync instead of writing "{}". Previously clear() left an empty {} on disk, and un-validated stores would return {} from load() instead of null, breaking the contract shared with InMemoryPreferencesStore. Consumers using a validator were unaffected in practice because empty objects already validated to null.
  • Harness: createLayeredPreferences().patch() now serializes calls through a promise queue. Previous implementation was a plain read-modify-write, so two concurrent patch() calls could both read the same stale workspace JSON and one update would be silently dropped. Concurrency tests cover both different-field and same-field writes.
  • Harness: createTogglePreferenceCommand and createEnumPreferenceCommand now persist before mutating runtime. If disk persistence fails, the runtime state is never touched and the command returns { success: false, message: "Failed to persist …" }. If the runtime set() throws after a successful persist, the disk write is rolled back to the previous value. This eliminates the class of "disk and runtime disagree" bugs across all factory-backed commands.
  • CEA: patchWorkspacePreferences now queues writes per-store via a WeakMap-keyed promise chain, matching the harness bundle semantics. This protects the shim path used by /reasoning-mode and /tool-fallback from the same lost-update race that previously affected the factory path.
  • CEA: applyPersistedPreferencesToAgentManager and applySharedConfigToAgentManager extracted from main.ts into packages/cea/src/entrypoints/preferences-startup.ts. Now pure functions taking the agent manager and store as arguments, so startup wiring is testable in isolation. Existing main.ts call sites updated to pass dependencies explicitly.
  • CEA: minimal-agent gains vitest devDep, test script, vitest.config.ts, and excludes *.test.ts / vitest.config.ts from the TypeScript build output. citty was a dead devDependency — removed.
  • Tests added:
    • Harness preferences-store.test.ts: clear() ⇒ load() null, clear() removes file on disk, clear() no-op on missing file, cross-instance restart, three concurrent-patch scenarios (different fields, same field, queue recovery after failure).
    • Harness preference-commands.test.ts: persist-first contract — toggle + enum commands do NOT mutate runtime when persistence fails, toggle command rolls back disk when runtime set throws, toggle command returns success only when BOTH disk and runtime succeed.
    • CEA user-preferences.test.ts: concurrent patchWorkspacePreferences preserves all fields, same-field is last-writer-wins, save errors bubble to caller.
    • CEA preferences-persistence.test.ts (new): singleton lifecycle, custom onError handler fires on save failure, concurrent persistPreferencePatch preserves all fields.
    • CEA reasoning-mode.test.ts (new): 7 tests covering status report, valid transitions, invalid mode rejection, persistence, sibling field preservation, no-op "already using", graceful behavior without persistence configured.
    • CEA tool-fallback.test.ts (new): 7 tests in the same shape.
    • CEA preferences-startup.test.ts (new): integration tests for applyPersistedPreferencesToAgentManager and applySharedConfigToAgentManager covering all three fields, CLI-vs-persisted precedence, and the invariant that CLI flags never write to the store.
    • minimal-agent preferences.test.ts (new): 7 tests covering path expectations, schema validation, round-trip, and the /reasoning onBeforeTurn integration (toggle → closure mutation → providerOptions change).
  • Test counts:
    • Harness: 673 → 699 (+26)
    • CEA: 526 → 563 (+37)
    • minimal-agent: 0 → 7 (+7)
    • Total workspace: 1283 → 1353 (+70)

• 496ffdb: Surface the "prompt processing" state that previously looked frozen, and fix follow-up correctness gaps found during post-implementation review.

  • Harness: new LoopHooks.onStreamStart / onFirstStreamPart hooks wrap the agent.stream() call site so consumers driving turns through runAgentLoop can react to the prompt-processing latency gap. onFirstStreamPart receives the current stream part as its first argument (TextStreamPart<ToolSet>) so consumers can inspect part.type to filter framing chunks (start, text-start, …) from visible content. TextStreamPart is re-exported from the harness root for convenience. Docstring clarifies that the TUI has its own independent onStreamStart on AgentTUIConfig.
  • TUI: shows a Processing... loader during turn preparation and transitions to Working... once the LLM request is in flight. The startup token probe is now non-blocking (fire-and-forget) so the editor accepts input immediately; the context-usage footer starts from the estimated count and quietly upgrades to the real value. During a blocking compaction the foreground loader temporarily switches to Compacting... and restores the previous label when the block ends, so users see the actual reason for a long wait. text-start stream parts are now treated as visible, clearing the streaming loader as soon as the assistant view mounts (no more empty-view flicker).
  • Headless: emits a turn-start lifecycle annotation and a matching onStreamStart callback before each LLM request; the event is dropped from trajectory.json (transient UX signal, no step_id) so persisted consumers see identical output. The event fires exactly once per logical turn — overflow and no-output retries no longer re-emit it. New tests cover normal ordering, new-turn vs intermediate-step phases, retry single-emission, and non-persistence in trajectory.json.
  • Headless: the persisted schema_version is corrected from the internal ATIF-v1.6 label to the actual current Harbor spec version ATIF-v1.4 (<https://www.harborframework.com/docs/agents/traject...

plugsuits@2.3.4

Patch Changes

• Updated dependencies [a714664]
  • @ai-sdk-tool/harness@1.3.0
  • @ai-sdk-tool/tui@3.1.0
  • @ai-sdk-tool/headless@3.1.0

@ai-sdk-tool/tui@3.1.0

Minor Changes

• a714664: Add defineAgent, createAgentRuntime, and AgentSession runtime layer to harness. Add runAgentSessionTUI and runAgentSessionHeadless session adapter helpers to tui and headless. Remove deprecated SessionStore, CheckpointHistory.fromSession(), and legacy token field aliases (completionTokens, promptTokens).

@ai-sdk-tool/headless@3.1.0

Minor Changes

• a714664: Add defineAgent, createAgentRuntime, and AgentSession runtime layer to harness. Add runAgentSessionTUI and runAgentSessionHeadless session adapter helpers to tui and headless. Remove deprecated SessionStore, CheckpointHistory.fromSession(), and legacy token field aliases (completionTokens, promptTokens).

@ai-sdk-tool/harness@1.3.0

Minor Changes

• a714664: Add defineAgent, createAgentRuntime, and AgentSession runtime layer to harness. Add runAgentSessionTUI and runAgentSessionHeadless session adapter helpers to tui and headless. Remove deprecated SessionStore, CheckpointHistory.fromSession(), and legacy token field aliases (completionTokens, promptTokens).

plugsuits@2.3.3

Patch Changes

• 5e0768c: Fix review issues: runAgentLoop message retention, isContextOverflowError call sites, setTimeout leak, CEA token estimation, session history separation, per-thread memory tracking, vi.mock hoisting, AgentError export, and lint cleanup
• Updated dependencies [5e0768c]
  • @ai-sdk-tool/harness@1.2.4
  • @ai-sdk-tool/tui@3.0.2
  • @ai-sdk-tool/headless@3.0.3