fix(docs): restore overrides/ directory for mkdocs Material theme

.agents/skills/sync-acp-spec/SKILL.md

@@ -74,9 +74,9 @@ Keep `src/acp/schema/` aligned with the official Agent Client Protocol reference
    - `agent/implementations/debug_server/mock_agent.py` — add stubs to `MockAgent`
 
 7. **Wire up in the ACP server** (if needed):
-   - `src/agentpool_server/acp_server/acp_agent.py` — implement new methods, advertise new capabilities
-   - `src/agentpool_server/acp_server/event_converter.py` — emit new session update types
-   - `src/agentpool_server/acp_server/session.py` — store new state
+   - `src/wolfharness_server/acp_server/acp_agent.py` — implement new methods, advertise new capabilities
+   - `src/wolfharness_server/acp_server/event_converter.py` — emit new session update types
+   - `src/wolfharness_server/acp_server/session.py` — store new state
 
 8. **Update the synced commit hash** in `src/acp/__init__.py`:
    ```bash
@@ -106,7 +106,7 @@ Keep `src/acp/schema/` aligned with the official Agent Client Protocol reference
 - `connection.py` (base JSON-RPC connection) — only if the wire protocol itself changes
 - `transports.py` — transport layer is our own
 - `stdio.py` — our process management layer
-- `client/` implementations (`DefaultACPClient`, `HeadlessACPClient`, `NoOpClient`) — agentpool-specific
+- `client/` implementations (`DefaultACPClient`, `HeadlessACPClient`, `NoOpClient`) — wolfharness-specific
 - `registry/` — our agent registry, not part of ACP spec
 - `task/` — our task abstraction
 - `tool_call_reporter.py`, `tool_call_state.py` — our extensions

.agents/skills/sync-skills-spec/SKILL.md

@@ -1,11 +1,11 @@
 ---
 name: sync-skills-spec
-description: Sync the agentpool skills implementation with the official Agent Skills Spec by comparing against the reference repository.
+description: Sync the wolfharness skills implementation with the official Agent Skills Spec by comparing against the reference repository.
 ---
 
 # Sync Skills Spec
 
-Keep `src/agentpool/skills/` aligned with the official Agent Skills Spec.
+Keep `src/wolfharness/skills/` aligned with the official Agent Skills Spec.
 
 ## Steps
 
@@ -21,18 +21,18 @@ Keep `src/agentpool/skills/` aligned with the official Agent Skills Spec.
    - Focus on the `SkillProperties` model, frontmatter fields, validation rules, and prompt generation
 
 3. **Compare with our implementation**:
-   - `src/agentpool/skills/skill.py` — our `Skill` Pydantic model, validators, `parse_frontmatter()`, `to_prompt()`
-   - `src/agentpool/skills/registry.py` — discovery and registration (our extension, not part of the spec)
-   - `src/agentpool/skills/manager.py` — pool-wide management (our extension)
+   - `src/wolfharness/skills/skill.py` — our `Skill` Pydantic model, validators, `parse_frontmatter()`, `to_prompt()`
+   - `src/wolfharness/skills/registry.py` — discovery and registration (our extension, not part of the spec)
+   - `src/wolfharness/skills/manager.py` — pool-wide management (our extension)
    - Check for missing fields, changed validation rules, new spec requirements
 
 4. **Apply updates** to our code:
    - Add any new frontmatter fields to the `Skill` model
    - Update validators to match spec changes (name format, length limits, allowed fields)
    - Update `to_prompt()` if the recommended XML format changed
-   - Do NOT replace our registry/manager — those are agentpool-specific extensions
+   - Do NOT replace our registry/manager — those are wolfharness-specific extensions
 
-5. **Update the synced commit hash** in `src/agentpool/skills/skill.py`:
+5. **Update the synced commit hash** in `src/wolfharness/skills/skill.py`:
    - Get the current HEAD of the cloned repo: `git -C "$tmp/agentskills" rev-parse HEAD`
    - Update the `SPEC_SYNCED_COMMIT` constant to the new hash
    - This tracks which version of the spec we last synced against

.github/workflows/build.yml

@@ -54,7 +54,7 @@ jobs:
         env:
           OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-        run: uv run --no-group docs pytest --cov-report=xml --cov=src/agentpool/ --cov-report=term-missing
+        run: uv run --no-group docs pytest --cov-report=xml --cov=src/wolfharness/ --cov-report=term-missing
 
       - name: VCR tests
         run: uv run pytest tests/vcr/ --strict-vcr-cassette-usage --tb=short --no-cov -q
@@ -146,7 +146,7 @@ jobs:
         run: uv sync --all-extras
 
       - name: Build with PyInstaller
-        run: uv run pyinstaller agentpool.spec
+        run: uv run pyinstaller wolfharness.spec
 
       - name: Get package version (Unix)
         id: get_version_unix
@@ -167,23 +167,23 @@ jobs:
         if: ${{ matrix.os != 'windows' }}
         run: |
           cd dist
-          zip agentpool-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_unix.outputs.version }}.zip agentpool
+          zip wolfharness-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_unix.outputs.version }}.zip wolfharness
 
       - name: Zip binary (Windows)
         if: ${{ matrix.os == 'windows' }}
         shell: pwsh
         run: |
           cd dist
-          Compress-Archive -Path agentpool.exe -DestinationPath agentpool-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_windows.outputs.version }}.zip
+          Compress-Archive -Path wolfharness.exe -DestinationPath wolfharness-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_windows.outputs.version }}.zip
 
       - name: Upload binary to release (Unix)
         if: ${{ matrix.os != 'windows' }}
         uses: softprops/action-gh-release@v2
         with:
-          files: dist/agentpool-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_unix.outputs.version }}.zip
+          files: dist/wolfharness-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_unix.outputs.version }}.zip
 
       - name: Upload binary to release (Windows)
         if: ${{ matrix.os == 'windows' }}
         uses: softprops/action-gh-release@v2
         with:
-          files: dist/agentpool-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_windows.outputs.version }}.zip
+          files: dist/wolfharness-${{ matrix.os }}-${{ matrix.arch }}-${{ steps.get_version_windows.outputs.version }}.zip

.github/workflows/publish_docker.yml

@@ -33,7 +33,7 @@ jobs:
         id: meta
         uses: docker/metadata-action@v5
         with:
-          images: ghcr.io/${{ github.repository }}/agentpool
+          images: ghcr.io/${{ github.repository }}/wolfharness
           tags: |
             type=semver,pattern={{version}}
             type=ref,event=branch

.github/workflows/pytest.yml

@@ -131,7 +131,7 @@ jobs:
         run: uv sync --all-extras --no-group lint
       - name: Import smoke test
         run: |
-          uv run python -c "mods=['agentpool','agentpool_config','agentpool_cli','agentpool_commands','agentpool_storage','agentpool_prompts','agentpool_server','agentpool_toolsets','acp'];[(__import__(m),print(f'  OK {m}')) for m in mods];print(f'All {len(mods)} modules imported successfully')"
+          uv run python -c "mods=['wolfharness','wolfharness_config','wolfharness_cli','wolfharness_commands','wolfharness_storage','wolfharness_prompts','wolfharness_server','wolfharness_toolsets','acp'];[(__import__(m),print(f'  OK {m}')) for m in mods];print(f'All {len(mods)} modules imported successfully')"
       - name: Test collection check
         run: uv run pytest --co -q 2>&1 | tail -3
 

.gitignore

@@ -66,6 +66,9 @@ test_configs/
 # Prometheus / OpenCode workspace files
 .omo/
 
+# Codegraph local index cache
+.codegraph/
+
 # Git worktrees
 .worktrees/
 

.opencode/agents/review-code.md

@@ -44,8 +44,8 @@ code changes in a pull request and report findings. You do NOT edit files.
    - No `asyncio.create_task()` without an active span
 
 5. **Architecture**
-   - Config models import from `agentpool_config.*`, not `agentpool.models`
-   - New modules follow existing patterns in `src/agentpool/`
+   - Config models import from `wolfharness_config.*`, not `wolfharness.models`
+   - New modules follow existing patterns in `src/wolfharness/`
    - New capabilities registered properly in the capability system
 
 ## Output Format

.opencode/agents/review-docs.md

@@ -34,15 +34,15 @@ documentation. You do NOT edit files.
 ## What to Check
 
 1. **Context Loading Table**
-   - New `src/agentpool/**/AGENTS.md` files MUST be registered in the root
+   - New `src/wolfharness/**/AGENTS.md` files MUST be registered in the root
      `AGENTS.md` Context Loading table.
    - Removed subsystem AGENTS.md files MUST be removed from the table.
 
 2. **Subsystem Documentation**
-   - New modules under `src/agentpool/` that introduce a new directory SHOULD have
+   - New modules under `src/wolfharness/` that introduce a new directory SHOULD have
      a corresponding `AGENTS.md` in that directory.
-   - Changes to `src/agentpool/lifecycle/`, `src/agentpool/capabilities/`,
-     `src/agentpool/skills/`, `src/agentpool/hooks/` MUST check whether the
+   - Changes to `src/wolfharness/lifecycle/`, `src/wolfharness/capabilities/`,
+     `src/wolfharness/skills/`, `src/wolfharness/hooks/` MUST check whether the
      subsystem AGENTS.md needs updating.
 
 3. **Explanation Docs**

.opencode/agents/review-lead.md

@@ -55,10 +55,10 @@ single review comment. You do NOT edit files or run commands.
 ## When to Invoke review-docs
 
 Always invoke `@review-docs` when the PR:
-- Adds or removes files under `src/agentpool/`
+- Adds or removes files under `src/wolfharness/`
 - Modifies any `AGENTS.md`
 - Modifies anything under `docs/`
-- Adds a new directory under `src/agentpool/` (potential new subsystem)
+- Adds a new directory under `src/wolfharness/` (potential new subsystem)
 - Changes code style conventions, testing rules, or telemetry rules
 - Adds or removes lifecycle dimensions, capabilities, protocols, or tools
 

.pre-commit-config.yaml

@@ -61,7 +61,7 @@ repos:
         name: check-jsonschema
         entry: uv run check-jsonschema --schemafile schema/config-schema.json
         language: system
-        files: ^src/agentpool/config_resources/.*\.yml$|^docs/examples/.*/config\.yml$
+        files: ^src/wolfharness/config_resources/.*\.yml$|^docs/examples/.*/config\.yml$
         require_serial: true
 
   # Phase 4: pre-push — heavier checks, only when pushing (~30-60s)

AGENTS.md

@@ -11,6 +11,28 @@ All significant changes go through OpenSpec: `/opsx:explore` → `/opsx:propose`
 - Location: `openspec/` (24 capability specs, 34 archived changes)
 - CLI: `openspec` v1.4+ — see `openspec/config.yaml`
 
+## Changelog
+
+Every release requires a changelog entry. See `changelog/README.md` for the full convention.
+
+### Release workflow
+
+1. Development changes accumulate in `changelog/unreleased/` — one detail file per logical change (`YYYY-MM-DD-short-slug.md`).
+2. At release time:
+   - Rename `unreleased/` to the version number (e.g. `2.10.0/`).
+   - Write `README.md` (summary linking each detail file) and `RELEASE.md` (GitHub Release body).
+   - Update `pyproject.toml` version.
+   - Commit **before** creating the git tag.
+   - Create a fresh empty `unreleased/` for the next cycle.
+
+### Detail file format
+
+```markdown
+# Title
+
+What changed and why. Use `##` sections once the entry covers more than one thing.
+```
+
 ## Quick Commands
 
 ```bash
@@ -22,8 +44,8 @@ uv run ruff check src/                  # Lint
 uv run ruff format src/                 # Format
 uv run --no-group docs mypy src/        # Type check
 duty lint                               # All checks
-agentpool run <name> "prompt"           # Run agent
-agentpool serve-acp config.yml          # Start ACP server
+wolfharness run <name> "prompt"           # Run agent
+wolfharness serve-acp config.yml          # Start ACP server
 ```
 
 ## Code Style — Red Lines
@@ -32,7 +54,7 @@ agentpool serve-acp config.yml          # Start ACP server
 - PEP 8 via Ruff. Google-style docstrings (no types in Args section).
 - Type hints required (`mypy --strict`). **Never** use `as any`, `@ts-ignore`, or type suppressions.
 - `from __future__ import annotations` for forward references.
-- Use `TYPE_CHECKING` blocks to avoid circular imports. Config models import from `agentpool_config.*`, not `agentpool.models`.
+- Use `TYPE_CHECKING` blocks to avoid circular imports. Config models import from `wolfharness_config.*`, not `wolfharness.models`.
 - No `getattr`/`hasattr` — provide full type safety.
 - NEVER use shortcuts or leave TODOs unless explicitly asked.
 
@@ -50,28 +72,28 @@ Instrument critical-path code (RunLoop, Turn, delegation, protocol entry points)
 
 ## Key Files
 
-- `src/agentpool/delegation/pool.py` — AgentPool orchestration
-- `src/agentpool/agents/agent.py` — Native agent implementation
-- `src/agentpool/messaging/messagenode.py` — MessageNode base abstraction
-- `src/agentpool/orchestrator/core.py` — EventBus, SessionController
-- `src/agentpool/orchestrator/run.py` — RunHandle (RunLoop) lifecycle
-- `src/agentpool/models/manifest.py` — Configuration schema
-- `src/agentpool/capabilities/` — Capability system (M3)
-- `src/agentpool/lifecycle/` — Lifecycle dimensions (M2)
+- `src/wolfharness/delegation/pool.py` — AgentPool orchestration
+- `src/wolfharness/agents/agent.py` — Native agent implementation
+- `src/wolfharness/messaging/messagenode.py` — MessageNode base abstraction
+- `src/wolfharness/orchestrator/core.py` — EventBus, SessionController
+- `src/wolfharness/orchestrator/run.py` — RunHandle (RunLoop) lifecycle
+- `src/wolfharness/models/manifest.py` — Configuration schema
+- `src/wolfharness/capabilities/` — Capability system (M3)
+- `src/wolfharness/lifecycle/` — Lifecycle dimensions (M2)
 
 ## Context Loading
 
 | Working on | Read this |
 |---|---|
-| RunLoop / Turn / EventBus | `src/agentpool/orchestrator/AGENTS.md` |
-| Lifecycle dimensions | `src/agentpool/lifecycle/AGENTS.md` |
-| Capabilities / tools | `src/agentpool/capabilities/AGENTS.md` |
-| Skills system | `src/agentpool/skills/AGENTS.md` |
-| Hooks system | `src/agentpool/hooks/AGENTS.md` |
+| RunLoop / Turn / EventBus | `src/wolfharness/orchestrator/AGENTS.md` |
+| Lifecycle dimensions | `src/wolfharness/lifecycle/AGENTS.md` |
+| Capabilities / tools | `src/wolfharness/capabilities/AGENTS.md` |
+| Skills system | `src/wolfharness/skills/AGENTS.md` |
+| Hooks system | `src/wolfharness/hooks/AGENTS.md` |
 | Tests | `tests/AGENTS.md` |
-| Protocol servers | `src/agentpool_server/AGENTS.md` |
+| Protocol servers | `src/wolfharness_server/AGENTS.md` |
 | ACP protocol | `src/acp/AGENTS.md` |
-| Core framework overview | `src/agentpool/AGENTS.md` |
+| Core framework overview | `src/wolfharness/AGENTS.md` |
 | Architecture deep-dives | `docs/explanation/` |
 | Module structure | `docs/explanation/module-structure.md` |
 | Graph architecture | `docs/explanation/graph-architecture.md` |
@@ -83,7 +105,7 @@ Instrument critical-path code (RunLoop, Turn, delegation, protocol entry points)
 | ToolDisplayCapability (tool rename + diff + rich display events) | `docs/explanation/tool-display-capability.md` |
 | Telemetry rules | `docs/explanation/telemetry.md` |
 | Usage examples | `docs/explanation/usage-examples.md` |
-| Extending AgentPool | `docs/explanation/extending-agentpool.md` |
+| Extending AgentPool | `docs/explanation/extending-wolfharness.md` |
 | Where to put new docs | `docs/meta/documentation-guide.md` |
 
 ## Tool Compatibility

Dockerfile

@@ -9,5 +9,5 @@ RUN pip install --no-cache-dir "."
 EXPOSE 8000
 
 # Run the server
-ENTRYPOINT ["agentpool", "serve-api"]
+ENTRYPOINT ["wolfharness", "serve-api"]
 CMD ["--auto-discover"]