Automatically enhance a coding prompt with quality requirements, codebase context, and tool recommendations.

Args: prompt: The original developer prompt task_type: Override auto-detection (generation|refactor|debug|review|test|planning|auto) context_level: How much context to gather (minimal|auto|comprehensive)

Returns: Enhanced prompt with quality requirements and tool recommendations

Analyze a prompt without enhancement, returning the detected intent, entities, and recommended approach.

Args: prompt: The developer prompt to analyze

Returns: Structured intent analysis

Retrieve quality standards for a language/framework combination.

Args: language: Programming language (typescript, python, etc.) framework: Optional framework (react, fastapi, etc.) category: Filter to specific category (security|architecture|style|all)

Returns: Quality standards for the specified language/framework

Suggest which MCP tools should be used for a given intent.

Args: intent_description: Description of what you're trying to do available_mcps: Comma-separated list of available MCPs (optional) discover_capabilities: If True, query actual MCP capabilities for recommended MCPs

Returns: Tool recommendations with priorities and reasons

Get a verification checklist for a specific task type.

Args: task_type: Type of task (generation|refactor|debug|review|test) touches_security: Whether the task involves security-sensitive code

Returns: Verification checklist appropriate for the task

Validate generated code against quality standards.

Args: code: The code to validate language: Programming language (python|typescript|javascript|rust|go|auto) check_security: Validate against security standards check_architecture: Validate against architecture standards check_style: Validate against language-specific style standards severity_threshold: Minimum severity to include in results (error|warning|info)

Returns: Validation results with pass/fail, score, violations, and summary

Validate a plan for implementation by a less capable model. Returns a quality score and list of issues that need fixing.

Args: plan: The plan text to validate target_model: Model that will implement (haiku|flash|cheap|capable) Cheaper models require stricter plan quality.

Returns: Quality scores, issues list, and ready_for_cheap_model flag