# Componecat

> Componecat is a software component catalog: a hierarchical, richly-typed registry of every system, service, and library an organization operates — with structured metadata, Git integration, hosted documentation, and endpoint definitions. It exposes the catalog to AI coding agents over a Model Context Protocol (MCP) server so agents can derive contextual understanding of an organization's software landscape.

Catalogs are private and organization-scoped, so the surfaces below require authentication. The MCP server is the primary way for an AI agent to consume catalog data.

## For AI agents

- [MCP server](https://app.componecat.ai/api/mcp): primary agent surface (OAuth 2.0 device flow). Search and browse the catalog, read entity documentation, traverse relationships, and inspect endpoints. Catalog search takes a query language — `field:value` clauses combined with AND/OR/NOT, parentheses, comparisons, and ranges — so a whole question fits in one call. Also serves Componecat's own product documentation, so an agent can learn how the platform works and guide a user through it. Exposes a tool catalog plus an entity resource browser.
- [MCP registry entry](https://github.com/componecat/mcp): the server is published in the official MCP registry as `com.shaftware/componecat`; this repo holds its public metadata, per-client install instructions, and a stdio bridge (`npx @shaftware/componecat-mcp`) for stdio-only clients.
- [Product documentation](https://docs.componecat.ai/app/): how Componecat works — concepts, guides, and reference. Public, no authentication required.
- [REST API](https://app.componecat.ai/api/rest/v1): programmatic API for the catalog, authenticated with a personal access token.
- [REST OpenAPI spec](https://app.componecat.ai/api/rest/v1/openapi.json): machine-readable schema for the REST API.

## About

Componecat maintains an accurate, richly-described, hierarchical map of every piece of software in an organization. Entities are grouped into a customizable type taxonomy (systems, components, resources, and more) with explicit relationships (hierarchy, dependencies) and ownership by teams. Documentation attaches to each entity as Markdown sections (optionally synced from Git) and external links, and is searchable. The platform continuously reconciles the catalog against the real state of the software landscape via Git, cloud, and CI/CD integrations.