
Understand Quickly
A public registry that serves pre-built code knowledge graphs to AI agents over HTTPS. Instead of parsing repositories on every request, agents fetch validated JSON graphs from a central index. The MCP exposes three tools: list_repos returns the catalog, get_graph fetches a specific graph by URL, and search_concepts filters entries by topic or format. Supports multiple graph formats including Understand-Anything, GitNexus, and generic bundle schemas. The registry validates every entry against JSON schemas and syncs nightly to catch stale links. Useful when you want LLMs to operate on structured codebase metadata without running indexers locally. Graphs stay in source repos; the registry only holds pointers, so there's no vendor lock-in or hosting cost.
- GitHub stars
- 34
- tools
- 3
- transport
- STDIO
- last updated
- May 10, 2026
Tools
3 toolsPublic tool metadata for what this MCP can expose to an agent.
list_repos
No parameter schema in public metadata yet.
get_graph
No parameter schema in public metadata yet.
search_concepts
No parameter schema in public metadata yet.
Sponsors
Readme
🧠understand-quickly
A public, machine-readable registry of code-knowledge graphs.
Point AI agents at any indexed repo and they get a current, schema-validated graph — one URL, one fetch.
Details
- Repository
- github.com/looptech-ai/understand-quickly
- Category
- Documents & Knowledge
- Package
@looptech-ai/understand-quickly-mcp- Transport
- STDIO
- Registry
- MCP registry, active