Initial commit

hooks/examples/context/code-review-start.md

@@ -0,0 +1,36 @@
+## Project-Specific Code Review Requirements
+
+This file demonstrates convention-based context injection.
+
+**Location:** `.claude/context/code-review-start.md`
+
+**Triggered by:** Running a code review command (SlashCommandStart hook)
+
+**Purpose:** Inject project-specific review requirements automatically.
+
+---
+
+### Additional Security Checks
+
+For this project, code reviews MUST verify:
+
+1. **Authentication:** All API endpoints require valid JWT
+2. **Input Validation:** All user inputs use allowlist validation
+3. **Rate Limiting:** Public endpoints have rate limits configured
+4. **Logging:** No PII in application logs
+
+### Performance Requirements
+
+- Database queries: No N+1 patterns
+- API response time: < 200ms for p95
+- Memory usage: No leaks detected in tests
+
+### Documentation
+
+- Public APIs have JSDoc/TSDoc comments
+- Complex algorithms have inline explanations
+- Breaking changes noted in CHANGELOG.md
+
+---
+
+**To use:** Copy to `.claude/context/code-review-start.md` in your project.

hooks/examples/context/plan-start.md

@@ -0,0 +1,32 @@
+## Project Planning Template
+
+**Location:** `.claude/context/plan-start.md`
+
+**Triggered by:** Running a planning command (SlashCommandStart hook)
+
+Your implementation plan must include:
+
+### Architecture Impact
+- Which services/modules are affected?
+- Any new dependencies introduced?
+- Database schema changes required?
+
+### API Surface
+- New endpoints or breaking changes?
+- Version bump needed?
+- Backward compatibility strategy?
+
+### Testing Strategy
+- Unit test coverage target (80%+)
+- Integration tests for new flows
+- E2E tests for user-facing features
+
+### Deployment Considerations
+- Feature flags required?
+- Migration scripts needed?
+- Rollback strategy?
+
+### Success Criteria
+- What does "done" look like?
+- How to verify it works?
+- What metrics to monitor?

hooks/examples/context/session-start.md

@@ -0,0 +1,41 @@
+# Session Start Context
+
+This file provides environment context at the beginning of each Claude Code session.
+
+## Plugin Environment
+
+**CLAUDE_PLUGIN_ROOT:** `${pwd}`
+
+This variable points to the root directory of the CipherPowers plugin installation.
+
+## Path Reference Convention
+
+When referencing plugin files in agents, commands, or skills, always use:
+
+```markdown
+@${CLAUDE_PLUGIN_ROOT}skills/skill-name/SKILL.md
+@${CLAUDE_PLUGIN_ROOT}standards/standard-name.md
+@${CLAUDE_PLUGIN_ROOT}principles/principle-name.md
+@${CLAUDE_PLUGIN_ROOT}templates/template-name.md
+```
+
+**Do NOT use relative paths without the variable:**
+```markdown
+@skills/...  ❌ Does not work in subagent contexts
+```
+
+## Usage
+
+Copy this file to your project's `.claude/context/` directory to inject plugin environment information at session start:
+
+```bash
+mkdir -p .claude/context
+cp ${CLAUDE_PLUGIN_ROOT}hooks/examples/context/session-start.md \
+   .claude/context/session-start.md
+```
+
+**Note:** SessionStart is not currently a supported hook in Claude Code. This file serves as a template for injecting environment context via other hooks (e.g., UserPromptSubmit, SlashCommandStart).
+
+## Alternative: User Prompt Hook
+
+If SessionStart hook becomes available, this context will auto-inject. Until then, consider using UserPromptSubmit hook or command-specific context injection.

hooks/examples/context/test-driven-development-start.md

@@ -0,0 +1,40 @@
+## Project TDD Standards
+
+**Location:** `.claude/context/test-driven-development-start.md`
+
+**Triggered by:** When `test-driven-development` skill loads (SkillStart hook)
+
+This project uses:
+
+- **Test framework:** Vitest
+- **Test location:** `src/**/__tests__/*.test.ts`
+- **Coverage requirement:** 80% line coverage minimum
+- **Property testing:** Use fast-check for algorithms
+
+### File Structure
+```
+src/
+  components/
+    Button/
+      Button.tsx
+      __tests__/
+        Button.test.tsx
+```
+
+### Naming Convention
+- Use `describe/it` blocks (not `test()`)
+- Test names: "should [behavior] when [condition]"
+- File naming: `{Component}.test.ts`
+
+### Mocking Strategy
+- Mock external services (APIs, databases)
+- Do NOT mock internal modules (test real behavior)
+- Use MSW for HTTP mocking
+
+### RED-GREEN-REFACTOR
+1. Write failing test first
+2. Run test (verify it fails for right reason)
+3. Write minimal code to pass
+4. Run test (verify it passes)
+5. Refactor (if needed)
+6. Commit

hooks/examples/convention-based.json

@@ -0,0 +1,26 @@
+{
+  "description": "Demonstrates convention-based context injection with explicit gates",
+  "comment": "Combines zero-config conventions with explicit verification gates",
+
+  "gates": {
+    "test": {
+      "description": "Run project test suite",
+      "comment": "Examples: 'npm test' | 'cargo test' | 'mise run test'",
+      "command": "npm test",
+      "on_pass": "CONTINUE",
+      "on_fail": "BLOCK"
+    }
+  },
+
+  "hooks": {
+    "SlashCommandEnd": {
+      "comment": "Convention file .claude/context/code-review-end.md auto-injects if exists",
+      "enabled_commands": ["/code-review"],
+      "gates": ["test"]
+    },
+    "SkillStart": {
+      "comment": "Convention file .claude/context/test-driven-development-start.md auto-injects",
+      "enabled_skills": ["test-driven-development"]
+    }
+  }
+}

hooks/examples/permissive.json

@@ -0,0 +1,31 @@
+{
+  "description": "Permissive mode - warn only, never block. Supports convention-based context injection.",
+  "comment": "Context files in .claude/context/ auto-inject without configuration. See CONVENTIONS.md",
+
+  "gates": {
+    "check": {
+      "description": "Quality checks (warn only)",
+      "comment": "Examples: 'npm run lint' | 'cargo clippy' | 'mise run check'",
+      "command": "mise run check",
+      "on_pass": "CONTINUE",
+      "on_fail": "CONTINUE"
+    },
+    "test": {
+      "description": "Tests (warn only)",
+      "comment": "Examples: 'npm test' | 'cargo test' | 'mise run test'",
+      "command": "mise run test",
+      "on_pass": "CONTINUE",
+      "on_fail": "CONTINUE"
+    }
+  },
+  "hooks": {
+    "PostToolUse": {
+      "enabled_tools": ["Edit", "Write"],
+      "gates": ["check"]
+    },
+    "SubagentStop": {
+      "enabled_agents": [],
+      "gates": ["check", "test"]
+    }
+  }
+}

hooks/examples/pipeline.json

@@ -0,0 +1,38 @@
+{
+  "gates": {
+    "format": {
+      "description": "Auto-format code",
+      "comment": "Examples: 'npm run format' | 'cargo fmt' | 'mise run format'",
+      "command": "mise run format",
+      "on_pass": "check",
+      "on_fail": "STOP"
+    },
+    "check": {
+      "description": "Quality checks",
+      "comment": "Examples: 'npm run lint' | 'cargo clippy' | 'mise run check'",
+      "command": "mise run check",
+      "on_pass": "test",
+      "on_fail": "BLOCK"
+    },
+    "test": {
+      "description": "Run tests",
+      "comment": "Examples: 'npm test' | 'cargo test' | 'mise run test'",
+      "command": "mise run test",
+      "on_pass": "build",
+      "on_fail": "BLOCK"
+    },
+    "build": {
+      "description": "Build project",
+      "comment": "Examples: 'npm run build' | 'cargo build' | 'mise run build'",
+      "command": "mise run build",
+      "on_pass": "CONTINUE",
+      "on_fail": "BLOCK"
+    }
+  },
+  "hooks": {
+    "SubagentStop": {
+      "enabled_agents": [],
+      "gates": ["format"]
+    }
+  }
+}

hooks/examples/strict.json

@@ -0,0 +1,38 @@
+{
+  "description": "Strict enforcement - block on all failures. Supports convention-based context injection.",
+  "comment": "Context files in .claude/context/ auto-inject without configuration. See CONVENTIONS.md",
+
+  "gates": {
+    "check": {
+      "description": "Quality checks must pass",
+      "comment": "Examples: 'npm run lint' | 'cargo clippy' | 'mise run check'",
+      "command": "mise run check",
+      "on_pass": "CONTINUE",
+      "on_fail": "BLOCK"
+    },
+    "test": {
+      "description": "All tests must pass",
+      "comment": "Examples: 'npm test' | 'cargo test' | 'mise run test'",
+      "command": "mise run test",
+      "on_pass": "CONTINUE",
+      "on_fail": "BLOCK"
+    },
+    "build": {
+      "description": "Build must succeed",
+      "comment": "Examples: 'npm run build' | 'cargo build' | 'mise run build'",
+      "command": "mise run build",
+      "on_pass": "CONTINUE",
+      "on_fail": "BLOCK"
+    }
+  },
+  "hooks": {
+    "PostToolUse": {
+      "enabled_tools": ["Edit", "Write", "mcp__serena__replace_symbol_body"],
+      "gates": ["check"]
+    },
+    "SubagentStop": {
+      "enabled_agents": [],
+      "gates": ["check", "test", "build"]
+    }
+  }
+}