← ClaudeAtlas

repo-intellisted

Native repository intelligence from git history, codegraph when indexed, manifest reads, and compressed repo digests; no cache, no generated state, every signal is recomputed on demand. Use when the user asks to analyze git history, find hotspots, code coupling, bus factor, bugspots, code ownership, entry points, or repo health.
OutlineDriven/odin-claude-plugin · ★ 27 · AI & Automation · score 85
Install: claude install-skill OutlineDriven/odin-claude-plugin
# repo-intel — native repository intelligence `extend` op-cell: add situational awareness before planning, review, onboarding, or risk triage. The skill answers temporal, social, graph, symbol, and project-shape questions with native tool recipes only; no prebuilt artifact is required. All signal recipes are in `references/signals.md`. Run the smallest recipe set that answers the user's query, then report the evidence shape directly: file, count, score, owner, last-seen, callers, callees, or manifest field. ## When to Apply - The user asks for git-history intelligence: hotspots, coldspots, bugspots, churn, recent activity, stale areas, file history. - The user asks for social risk: ownership, bus factor, stale owners, contributor concentration. - The user asks for code coupling: files that co-change, risky refactor surfaces, hidden module boundaries. - The user asks for symbol intelligence: entry points, exported symbols, callers, callees, dependents, impact. - The user asks for repo health: activity, high-churn/bug-fix overlap, owner staleness, entry-point clarity, manifest/tooling shape. - The user is about to refactor, review, onboard, or choose a next area and needs evidence before touching code. ## When NOT to Apply - The ask is a single known symbol lookup and the answer is already in context. - The ask needs runtime profiling, production telemetry, or benchmark evidence rather than repository-history evidence. - The directory is not a git repository and the user