marketplace-ai

Record browser-based feature demos as GIFs via Playwright and optionally document them in docs/features/index.md. Trigger with "Demo <feature>" or "Demo and Document <feature>".

Use when creating service modules in apps/api/skillhub/services/

Use when writing API tests in apps/api/tests/

Use when implementing division-based access control on any endpoint

Use when adding services to docker-compose.yml or creating Dockerfiles

Use when implementing error handling across the API, frontend, or MCP server

Use when creating new Flask blueprints in apps/api/skillhub_flask/blueprints/

Use when generating or managing the OpenAPI spec and TypeScript types

Use when generating or managing TypeScript types in libs/shared-types/

Use when reviewing or implementing accessibility in the React frontend

Use when building admin endpoints requiring platform_team or security_team roles

Use when creating or managing database migrations with Alembic

Use when implementing pagination for API endpoints

Use when making API changes that affect versioning or backward compatibility

Use when adding audit log entries for any mutation or security event

Create, regenerate, and maintain codebase maps as Tier 1 project context files. Use when "create a codebase map", "regenerate the codebase map", "sync the codebase map", "update my project's navigation index", or need a living navigational index to help agents quickly find code. Supports smart detection of tech stack and domain boundaries, socratic probing for discrepancies, user editing with intelligent sync, and post-creation updates to CLAUDE.md and related project docs. Handles monorepos, microservices, traditional single-repo layouts, and language-agnostic projects.

Use when building or extending the comment/reply system

Use when building shared React components in libs/ui/ or apps/web/src/components/

Council of Experts — dispatch parallel subagents as expert roles to debate, discuss, or collaboratively improve something. Use when '/council' is invoked. Supports natural language input to infer mode, agent count, specific skills, and goal.

Use when writing data migrations (backfills, counter updates) with Alembic

Use when optimizing SQLAlchemy queries or addressing N+1 problems

Use when managing environment variables and configuration

Use when adding new feature flags or gating features behind flags

Use when working with the skill fork system

Use when adding CI pipeline stages to .gitlab-ci.yml

Use when implementing install/uninstall tracking

Use when integrating LLM calls via LiteLLM/Bedrock for evaluation or content generation

Use when adding new MCP tools in apps/mcp-server/skillhub_mcp/tools/

Use when adding mise tasks in mise.toml

Use when implementing structured logging, health checks, or observability

Use when configuring NX monorepo projects and dependency rules

Use when implementing OAuth provider flows for production auth

Use when optimizing React rendering, API response times, or bundle size

Use when adding pre-commit hooks to .pre-commit-config.yaml

Use when creating Pydantic v2 schemas in apps/api/skillhub/schemas/

Use when building shared Python utilities in libs/python-common/

Use when adding quality checks to the CI/local quality gate

Use when creating custom React hooks in apps/web/src/hooks/

Use when creating new React views in apps/web/src/views/

Comprehensive repository visualization, analysis, and observatory for Benefitly. Use when "visualize the repo", "show me bottlenecks", "where are test gaps", "what are the scaling concerns", "generate repo report", "update observatory", "show dependency graph", "identify antipatterns", "dead code scan", "repo health dashboard", "infrastructure topology", "concern map", "action plan for improvements", "what needs attention", "repo overview", or any request to analyze, visualize, or report on codebase health, architecture, test coverage, data flows, or areas of concern. Operates in three modes: dashboard (full scan), analyst (targeted query), observatory (persistent docs). Orchestrates via subagent dispatch to codebase-health-check, codebase-map-generator, observability-master, and other existing skills. Produces Mermaid diagrams, scored reports, and action plans with file:line citations.