moai-adk 0.8.0__py3-none-any.whl → 0.34.0__py3-none-any.whl

moai_adk/templates/.claude/skills/moai-foundation-core/reference.md

@@ -0,0 +1,478 @@
+# moai-foundation-core Reference
+
+Progressive Disclosure Level 2: Extended documentation for foundational principles and architectural patterns.
+
+---
+
+## API Reference
+
+### Core Concepts
+
+TRUST 5 Framework:
+- Purpose: Automated quality gate system for code quality assurance
+- Components: Test-first, Readable, Unified, Secured, Trackable
+- Integration: Pre-commit hooks, CI/CD pipelines, agent workflows
+- Validation: Automated checks with configurable thresholds
+
+SPEC-First TDD:
+- Purpose: Specification-driven development workflow
+- Phases: SPEC (Plan), TDD (Run), Docs (Sync)
+- Format: EARS (Event-Action-Response-State) specifications
+- Token Budget: 30K (SPEC) + 180K (TDD) + 40K (Docs) = 250K total
+
+Delegation Patterns:
+- Purpose: Task orchestration via specialized agents
+- Core Principle: Alfred delegates all work through Task() calls
+- Patterns: Sequential, Parallel, Conditional delegation
+- Agent Selection: Complexity-based agent matching
+
+Token Optimization:
+- Purpose: Efficient 200K token budget management
+- Strategies: Phase separation, selective loading, context reset
+- Savings: /clear after phases, model selection optimization
+- Monitoring: /context command for budget tracking
+
+Progressive Disclosure:
+- Purpose: Three-tier knowledge delivery architecture
+- Levels: Quick (30s), Implementation (5min), Advanced (10+min)
+- Token Allocation: 1K, 3K, 5K tokens respectively
+- Structure: SKILL.md core with modules/ overflow
+
+Modular System:
+- Purpose: Scalable file organization for unlimited content
+- Structure: SKILL.md (500 lines) + modules/ (unlimited)
+- Principles: Topic-focused, self-contained modules
+- Discovery: Progressive navigation from core to deep dive
+
+---
+
+## Configuration Options
+
+### TRUST 5 Thresholds
+
+```yaml
+trust_5:
+  test_first:
+    min_coverage: 0.85           # Minimum 85% test coverage
+    coverage_tool: "pytest-cov"  # Coverage measurement tool
+    failure_action: "block"      # block, warn, ignore
+
+  readable:
+    linter: "ruff"               # Code linter tool
+    max_complexity: 10           # Maximum cyclomatic complexity
+    naming_convention: "snake_case"
+    failure_action: "warn"
+
+  unified:
+    formatter: "black"           # Code formatter
+    import_sorter: "isort"       # Import organization
+    line_length: 88              # Maximum line length
+    failure_action: "auto_fix"
+
+  secured:
+    scanner: "bandit"            # Security scanner
+    owasp_compliance: true       # OWASP Top 10 check
+    secret_detection: true       # Credential scanning
+    failure_action: "block"
+
+  trackable:
+    commit_format: "conventional"  # Commit message format
+    branch_naming: "feature/*"     # Branch naming pattern
+    changelog_required: true       # Changelog requirement
+    failure_action: "warn"
+```
+
+### Token Budget Allocation
+
+```yaml
+token_budget:
+  total: 250000                  # Total available tokens
+
+  phases:
+    spec:
+      budget: 30000              # 30K for SPEC phase
+      strategy: "minimal_context"
+      clear_after: true          # Execute /clear after phase
+
+    tdd:
+      budget: 180000             # 180K for TDD phase
+      strategy: "selective_loading"
+      file_priority: ["tests", "src", "config"]
+
+    docs:
+      budget: 40000              # 40K for docs phase
+      strategy: "cached_results"
+      template_reuse: true
+
+  optimization:
+    clear_threshold: 150000      # Suggest /clear at this level
+    message_limit: 50            # Suggest /clear after messages
+    model_selection:
+      quality: "sonnet"          # For quality-critical tasks
+      speed: "haiku"             # For speed/cost optimization
+```
+
+### Agent Selection Matrix
+
+```yaml
+agent_selection:
+  by_complexity:
+    simple:                      # 1 file
+      agents: 1-2
+      pattern: "sequential"
+
+    medium:                      # 3-5 files
+      agents: 2-3
+      pattern: "sequential"
+
+    complex:                     # 10+ files
+      agents: 5+
+      pattern: "mixed"
+
+  by_domain:
+    backend:
+      primary: "backend-expert"
+      supporting: ["database-expert", "api-designer"]
+
+    frontend:
+      primary: "frontend-expert"
+      supporting: ["ui-specialist", "accessibility-checker"]
+
+    security:
+      primary: "security-expert"
+      supporting: ["compliance-checker", "penetration-tester"]
+
+    infrastructure:
+      primary: "devops-expert"
+      supporting: ["cloud-architect", "monitoring-specialist"]
+```
+
+---
+
+## Integration Patterns
+
+### Pattern 1: TRUST 5 Pre-Commit Hook
+
+```python
+# Pre-commit hook integration for TRUST 5 validation
+from moai_foundation_core import TRUST5Validator
+
+def pre_commit_validation(staged_files: List[str]) -> bool:
+    """Run TRUST 5 validation on staged files."""
+
+    validator = TRUST5Validator()
+
+    # Run all pillars
+    results = {
+        "test_first": validator.check_test_coverage(staged_files),
+        "readable": validator.check_readability(staged_files),
+        "unified": validator.check_formatting(staged_files),
+        "secured": validator.check_security(staged_files),
+        "trackable": validator.check_commit_message()
+    }
+
+    # Evaluate results
+    all_passed = all(r.passed for r in results.values())
+
+    if not all_passed:
+        for pillar, result in results.items():
+            if not result.passed:
+                print(f"[TRUST 5] {pillar.upper()}: {result.message}")
+
+    return all_passed
+```
+
+### Pattern 2: SPEC-First TDD Workflow
+
+```python
+# Complete SPEC-First TDD cycle implementation
+from moai_foundation_core import SPECManager, TDDExecutor, DocsGenerator
+
+async def spec_first_workflow(requirements: str):
+    """Execute complete SPEC-First TDD workflow."""
+
+    # Phase 1: SPEC Generation (30K tokens)
+    spec_manager = SPECManager()
+    spec = await spec_manager.generate_spec(
+        requirements=requirements,
+        format="EARS"
+    )
+    print(f"Generated: {spec.id}")
+
+    # Critical: Clear context before Phase 2
+    await execute_clear()  # Saves 45-50K tokens
+
+    # Phase 2: TDD Implementation (180K tokens)
+    tdd = TDDExecutor(spec.id)
+
+    # RED: Create failing tests
+    tests = await tdd.generate_tests()
+    assert await tdd.run_tests() == "FAIL"
+
+    # GREEN: Implement to pass
+    implementation = await tdd.implement()
+    assert await tdd.run_tests() == "PASS"
+
+    # REFACTOR: Optimize
+    refactored = await tdd.refactor()
+    assert await tdd.run_tests() == "PASS"
+
+    # Validate coverage
+    coverage = await tdd.get_coverage()
+    assert coverage >= 0.85
+
+    # Phase 3: Documentation (40K tokens)
+    docs = DocsGenerator(spec.id)
+    await docs.generate_api_docs()
+    await docs.generate_architecture_diagrams()
+    await docs.update_project_readme()
+
+    return {"spec": spec, "coverage": coverage}
+```
+
+### Pattern 3: Intelligent Agent Delegation
+
+```python
+# Complexity-based agent delegation
+from moai_foundation_core import TaskAnalyzer, AgentRouter
+
+async def delegate_task(task_description: str, context: dict):
+    """Delegate task to appropriate agents based on complexity."""
+
+    # Analyze task complexity
+    analyzer = TaskAnalyzer()
+    analysis = analyzer.analyze(task_description, context)
+
+    print(f"Complexity: {analysis.complexity}")
+    print(f"Affected files: {analysis.file_count}")
+    print(f"Domains: {analysis.domains}")
+
+    # Route to appropriate agents
+    router = AgentRouter()
+
+    if analysis.complexity == "simple":
+        # Sequential single agent
+        result = await Task(
+            subagent_type=router.get_primary_agent(analysis.domains[0]),
+            prompt=task_description,
+            context=context
+        )
+
+    elif analysis.complexity == "medium":
+        # Sequential multiple agents
+        results = []
+        for domain in analysis.domains:
+            result = await Task(
+                subagent_type=router.get_primary_agent(domain),
+                prompt=f"Handle {domain} aspects: {task_description}",
+                context={**context, "previous_results": results}
+            )
+            results.append(result)
+
+    else:  # complex
+        # Parallel then sequential integration
+        parallel_results = await Promise.all([
+            Task(subagent_type=router.get_primary_agent(d), prompt=f"{d}: {task_description}")
+            for d in analysis.domains
+        ])
+
+        # Integration phase
+        result = await Task(
+            subagent_type="integration-specialist",
+            prompt="Integrate all components",
+            context={"results": parallel_results}
+        )
+
+    return result
+```
+
+### Pattern 4: Token-Optimized Context Management
+
+```python
+# Proactive token budget management
+from moai_foundation_core import TokenMonitor, ContextOptimizer
+
+class TokenAwareWorkflow:
+    """Workflow with automatic token optimization."""
+
+    def __init__(self, total_budget: int = 200000):
+        self.monitor = TokenMonitor(total_budget)
+        self.optimizer = ContextOptimizer()
+
+    async def execute_with_optimization(self, tasks: List[dict]):
+        """Execute tasks with automatic context management."""
+
+        results = []
+
+        for task in tasks:
+            # Check current usage
+            usage = self.monitor.get_current_usage()
+
+            if usage > 0.75:  # 75% threshold
+                print(f"Token usage: {usage*100:.1f}%")
+
+                # Apply optimization strategies
+                if usage > 0.9:
+                    print("Executing /clear to reset context")
+                    await execute_clear()
+                    self.monitor.reset()
+                else:
+                    # Selective context pruning
+                    self.optimizer.prune_least_relevant()
+
+            # Execute task
+            result = await self.execute_task(task)
+            results.append(result)
+
+            # Update token tracking
+            self.monitor.add_usage(result.tokens_used)
+
+        return results
+```
+
+---
+
+## Troubleshooting
+
+### Common Issues
+
+Issue: TRUST 5 validation fails inconsistently:
+- Cause: Different tool versions between local and CI
+- Solution: Pin tool versions in requirements.txt/pyproject.toml
+- Prevention: Use locked dependency files
+
+Issue: SPEC phase consumes too many tokens:
+- Cause: Excessive context loading during spec generation
+- Solution: Use minimal context mode for SPEC phase
+- Prevention: Configure `spec.strategy: "minimal_context"`
+
+Issue: Agent delegation creates infinite loops:
+- Cause: Circular dependencies between agents
+- Solution: Implement task ID tracking and loop detection
+- Prevention: Design clear agent responsibilities with no overlap
+
+Issue: Token budget exhausted mid-workflow:
+- Cause: Missing /clear between phases
+- Solution: Always execute /clear after Phase 1
+- Prevention: Enable automatic phase boundary detection
+
+Issue: Progressive disclosure not loading modules:
+- Cause: Incorrect cross-reference paths
+- Solution: Verify paths in SKILL.md match actual file locations
+- Prevention: Use relative paths from SKILL.md location
+
+### Diagnostic Commands
+
+```bash
+# TRUST 5 status check
+moai-trust check --all --verbose
+
+# Token budget analysis
+moai-context analyze --show-breakdown
+
+# Agent routing debug
+moai-agent route --task "description" --dry-run
+
+# SPEC validation
+moai-spec validate SPEC-001 --format EARS
+```
+
+### Validation Utilities
+
+```python
+from moai_foundation_core import diagnose
+
+# Full system diagnostics
+report = diagnose.run_full_check()
+
+# Individual component checks
+diagnose.check_trust5_tools()
+diagnose.check_agent_availability()
+diagnose.check_token_tracking()
+diagnose.verify_module_structure()
+```
+
+### Log Locations
+
+- TRUST 5 logs: `.moai/logs/trust5/`
+- SPEC artifacts: `.moai/specs/`
+- Agent execution logs: `.moai/logs/agents/`
+- Token usage history: `.moai/logs/tokens/`
+
+---
+
+## External Resources
+
+### Official Documentation
+
+- MoAI-ADK Documentation: See project README
+- Claude Code Skills Guide: https://docs.anthropic.com/claude-code/skills
+- EARS Specification Format: See `modules/spec-first-tdd.md`
+
+### Module References
+
+- TRUST 5 Framework: `modules/trust-5-framework.md`
+- SPEC-First TDD: `modules/spec-first-tdd.md`
+- Delegation Patterns: `modules/delegation-patterns.md`
+- Token Optimization: `modules/token-optimization.md`
+- Progressive Disclosure: `modules/progressive-disclosure.md`
+- Modular System: `modules/modular-system.md`
+- Agents Reference: `modules/agents-reference.md`
+- Commands Reference: `modules/commands-reference.md`
+- Execution Rules: `modules/execution-rules.md`
+
+### Related Skills
+
+- moai-foundation-claude - Claude Code integration patterns
+- moai-workflow-project - Project management with core principles
+- moai-workflow-testing - Testing workflows with TRUST 5
+- moai-workflow-templates - Template management integration
+
+### Tool References
+
+- pytest: https://docs.pytest.org/
+- ruff: https://docs.astral.sh/ruff/
+- black: https://black.readthedocs.io/
+- bandit: https://bandit.readthedocs.io/
+
+### Best Practices
+
+TRUST 5 Implementation:
+- Run all validations before commits
+- Configure appropriate failure actions
+- Document exceptions and waivers
+- Track quality metrics over time
+
+SPEC-First Development:
+- Write clear EARS specifications
+- Execute /clear between phases
+- Maintain 85%+ test coverage
+- Generate documentation with each feature
+
+Agent Delegation:
+- Match complexity to agent count
+- Use parallel execution for independent tasks
+- Implement proper error handling
+- Track delegation outcomes
+
+Token Management:
+- Monitor usage proactively
+- Clear context at phase boundaries
+- Use selective file loading
+- Choose appropriate model for task
+
+### Version History
+
+| Version | Date       | Changes                                           |
+|---------|------------|---------------------------------------------------|
+| 2.3.0   | 2025-12-03 | Added agents, commands, execution rules modules   |
+| 2.2.0   | 2025-11-28 | Optimized to 500-line SKILL.md                    |
+| 2.1.0   | 2025-11-25 | Added modular architecture patterns               |
+| 2.0.0   | 2025-11-20 | Unified six foundational principles               |
+| 1.0.0   | 2025-11-15 | Initial release with TRUST 5 and SPEC-First       |
+
+---
+
+Status: Reference Documentation Complete
+Last Updated: 2025-12-06
+Skill Version: 2.3.0