Enables AI agents and editors to run complexity inspections on the fly, returning compact offense JSON with rule, location, score, and vector data for any given path or inline source so oversized functions, bloated modules, and dead or single-use locals can be fixed in the same turn instead of surfacing later in CI.
Enables agents to run static heuristic scans over a repository path via stdio, surfacing review candidates with file:line evidence for web3 auth and payment vulnerability patterns. Results can be used as CI-style gates or triage input without any dataflow analysis or external dependencies.
Enables AI coding agents and other MCP clients to interrogate a repository — or a frontend and backend pair — returning a single JSON document of findings such as which calls reach which routes, what is dead, and where to refactor first. Exposes tools like analyze_repo, cross_repo, check_endpoint, and check_coverage, along with contract resources for authoring adapters and rule packs, entirely over local stdio with no network access.
Enables AI assistants to audit and clean development workspaces safely by detecting reclaimable artifacts, exposed .env files, and potential secret leaks.