{"repo":"proofofwork-agency/reporecall","free":true,"listed":false,"github":"https://github.com/proofofwork-agency/reporecall","clone":"git clone https://github.com/proofofwork-agency/reporecall.git","description":"Local-first, auto-injecting context + memory layer for Claude Code, Codex and MCP. Per-prompt hook injection, explicit freshness/Trust Contract, AST indexing, call graphs, hybrid search, wiki + architecture dashboard. Everything stays on your machine.","language":"TypeScript","stars":37,"topics":["ai-coding","ast","call-graph","claude-code","code-intelligence","code-navigation","code-search","codebase-search","developer-tools","mcp"],"license":"MIT","category":"mcp-servers","readme_excerpt":"Reporecall Local-first context + memory for coding agents Automatic injection + an explicit Trust Contract so agents (and you) always know when the context is fresh. Reporecall helps Claude Code, Codex, Cline, Aider and other agents work more effectively on large, high-churn or unfamiliar codebases . What you actually get - Auto-injection via hooks — relevant code, call graph, wiki, memory and business context are pushed into every prompt before the model starts thinking. - Brutally honest freshness — every response includes indexedCommit , dirty-file count, and a clear banner when the index is EMPTY or STALE, plus direct guidance to refresh context . - Smart local retrieval — intent routing + hybrid search + graph expansion + extractive compression (expand any chunk on demand). - Persistent local memory — rules, facts and working notes that survive across sessions. - Zero cloud by default — no external vector DB, no recurring costs, works completely offline. Measured, not asserted Every number below is produced by a command you can run, backed by a committed artifact, and registered in quality/claims.json . Anything we haven't measured says insufficient evidence instead of guessing. --- --- Getting the answering evidence in front of the model costs a median of 75.4% fewer tokens than reading the relevant files whole npm run benchmark:context-cost Retrieval holds 91.6% context precision and 95.6% recall at 8.4% pollution, with 100% route accuracy and zero high-confidence-wron","default_branch":null,"files":null,"tree":[],"storefront":"/r/proofofwork-agency","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/proofofwork-agency/reporecall/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}