---

name: graphify description: "Use Graphify to build, query, export, and integrate code/document knowledge graphs for AI coding assistants and MCP workflows." disable-model-invocation: true metadata: disco-role: operating license: NOASSERTION

Graphify

Use this repo skill when a user names Graphify or graphifyy, works with graphify-out/, wants a code/document knowledge graph, asks how to query graph.json, installs Graphify into an AI coding assistant, or needs Graphify exports, hooks, MCP serving, or extractor troubleshooting.

Graphify's public distribution is graphifyy; the import package and CLI are graphify. The normal graph output directory is graphify-out/ with graph.json, GRAPH_REPORT.md, and optional graph.html, wiki, export, cache, manifest, and lesson artifacts.

First checks

# Public install identity.
python -m pip show graphifyy || true
python - <<'PY'
import graphify
print('graphify import ok')
PY

# CLI check. Use python -m graphify when the console command is not on PATH.
graphify --help || python -m graphify --help

For a self-contained preflight that does not touch a repository, run scripts/check_graphify_install.py. Read references/package-overview.md for package concepts, graph artifacts, and optional surfaces.

Route map

User intent	Read
Build, update, watch, or validate graphify-out/ from local code/docs/media; choose --code-only vs semantic extraction; handle .graphifyignore, cache, manifest, cluster-only, or add	graph-building
Answer from an existing graph with query, path, explain, affected, god-nodes, work-memory lessons, or MCP stdio/HTTP serving	query-navigation
Install/uninstall Graphify assistant skills, always-on guidance, platform hooks, Git hooks, strict mode, or inspect platform artifacts safely	agent-integration
Export graph artifacts to HTML/wiki/Obsidian/SVG/GraphML/Cypher, merge multiple graphs, clone GitHub repos, use PR helpers, or push to Neo4j/FalkorDB only after approval	exports-integrations
Diagnose ignored file types, missing optional parsers, AST extractor/resolver failures, node-ID/source-file normalization, zero-node sources, or maintainer extractor changes	extractor-troubleshooting

If a request spans several routes, preserve the natural order: build/update the graph, then query it, then export or install always-on guidance as needed.

Cross-cutting references

• references/cli-command-map.md: command families and their owning sub-skills.
• references/troubleshooting.md: package/PATH, missing graph, optional extras, no-key, stale graph, service, and safety triage.
• references/repo-provenance.md: source snapshot and refresh baseline for this generated skill.
• references/repo-routing-metadata.json: structured router metadata for a later managed import.

Operating defaults

• Prefer graphify query, graphify path, or graphify explain before grepping raw files when a fresh graphify-out/graph.json exists.
• Use graphify extract <path> --code-only for local code graphs that should not require provider keys or semantic extraction.
• Use explicit optional extras only when the selected workflow needs them; do not install graphifyy[all] by default.
• Treat database pushes, provider API calls, Google Workspace auth, video/model downloads, GitHub PR triage, and HTTP MCP exposure beyond localhost as side-effecting or credentialed operations that need explicit user approval.
• Do not delete, purge, or force-overwrite graphify-out/ artifacts just because they are dirty or stale. Prefer update/rebuild routes and explain the overwrite risk first.
• Do not rely on the original Graphify source checkout. This runtime skill bundles the needed references and scripts.

Staleness check

Read references/repo-provenance.md before deciding whether this skill matches a newer Graphify checkout. If the package version, public CLI entry points, optional extras, or source evidence paths changed, refresh this repo skill before making detailed claims.