← ClaudeAtlas

vibe-graphlisted

Dual-layer dependency graph for vibe-* projects. No external packages. DEPENDENCY_GRAPH.json maps file-level relationships with confidence tagging: EXTRACTED (verified from source, confidence 1.0), INFERRED (derived from specs/patterns, 0.6–0.95), AMBIGUOUS (flagged for review). Rationale nodes capture WHY, HACK, DECISION comments from source files — surfaced during bug diagnosis and review to prevent intent drift. God nodes identify highest-coupling files — read first in review, flagged in bug diagnosis, blocked from simultaneous parallel writes. CONCEPT_GRAPH.json maps semantic relationships (features, agents, models, components). graph.html is an interactive visual. Auto-updated via git diff. Queried by vibe-fix-bug, vibe-test, vibe-review, vibe-parallel. Primary value is semantic navigation: accurate blast-radius, intent-drift detection via rationale nodes, and safe parallel dispatch via god nodes — not raw token savings (current models have large context windows + prompt caching, so the graph is a precis
aakashdhar/vibe-skill · ★ 7 · Code & Development · score 68
Install: claude install-skill aakashdhar/vibe-skill
# Vibe Graph Skill Dual-layer dependency graph for vibe-* projects. Starts as a spec graph before code exists. Grows into a full dependency map as the project is built. Updated automatically from git diff — no agent self-reporting. Three concepts from graphify integrated throughout: 1. **Relationship confidence tagging** — EXTRACTED / INFERRED / AMBIGUOUS on every edge 2. **Rationale nodes** — WHY, HACK, DECISION comments extracted from source files 3. **God nodes** — highest-degree files surfaced for priority review and safe parallel dispatch --- ## Execution model — extract with judgment, compute with the script Split the work so the model never hand-simulates graph algorithms (a real silent-error risk on any non-trivial codebase): - **Extraction is the LLM's job** (needs code comprehension): read the code / specs and write `vibe/graph/DEPENDENCY_GRAPH.json` and `CONCEPT_GRAPH.json` — the nodes, edges, confidence tags, and rationale. - **Computation is `scripts/graph.py`'s job** (deterministic): god-node degree ranking, blast-radius queries, stats, and the `graph.html` render. **Run it via Bash — do not mentally sort degrees or traverse the graph.** The Python shown in the reference files documents the algorithm; the script is what actually runs it. Resolve the script path across install layouts: ```bash GRAPH=$(ls ~/.claude/skills/vibe-graph/scripts/graph.py \ ~/.claude/plugins/marketplaces/*/skills/vibe-graph/scripts/graph.py 2>/dev/null | head