Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/rexhung0302/codegraph/claude-mdgit clone --depth 1 https://github.com/RexHung0302/codegraphWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00987 | $0.00987 |
| Opus 5 | $0.00494 | $0.00494 |
| Sonnet 5 | $0.00197 | $0.00197 |
| Haiku 4.5 | $0.00099 | $0.00099 |
Grade A, and why
codegraph CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Guidance for Claude Code when working in this repository.
What this is
codegraph extracts a TS/JS/Vue repo into a node/edge graph (graph.json), then serves it two ways: a WebGL viewer for humans, an MCP server for agents. There is no database and no backend — the graph is one JSON file passed between three independent pieces.
Layout
| Path | Role |
|---|---|
src/extract.ts |
CLI extractor. ts-morph AST walk → graph.json. The only place that knows about source languages. |
src/mcp.ts |
MCP server over stdio. Reads graph.json, never parses source. |
web/src/main.ts |
Viewer. graphology + sigma + ForceAtlas2 worker. |
web/public/graph.json |
Default extractor output. Git-ignored — regenerate, never commit. |
docs/graph-schema.md |
The contract between the three pieces. Update it when the shape changes. |
tests/ |
vitest suite; tests/fixtures/ are tiny repos the extractor runs against. |
Commands
pnpm extract <repo> [--out f] [--exclude a,b] [--types] # produce a graph
pnpm dev # viewer on :5173
pnpm mcp [graph.json] # MCP server on stdio
pnpm test # vitest, 52 cases
npx tsc --noEmit
There is no linter config. pnpm test and npx tsc --noEmit must both pass before any change is considered done.
tests/fixtures/ holds three miniature repos — a React app with tsconfig paths, a Vue CLI style app with no tsconfig and a pug template, and a types-app whose re-export and same-name method exist only to pin --types behaviour. They exist to pin behaviour that is easy to break silently: alias resolution, SFC parsing, renders edges, edge dedup, determinism. When you change extraction, add a fixture case rather than adjusting an assertion to match new output.
How to verify a change
Type-checking is not enough — extraction bugs type-check fine and silently produce a wrong graph.
- Extractor changes: re-run
pnpm extracton at least one React repo and one Vue repo (and, when touching resolution, once more with--types: it must only add edges, never remove any), and compare the per-type edge counts printed at the end against the previous run. A change that dropsimportsorrendersto near zero is the classic regression; that is how the alias and Vue bugs were caught. - MCP changes: drive the server with a real MCP client over stdio (
@modelcontextprotocol/sdkClient+StdioClientTransport) and call the affected tool. Do not assert from reading the code. - Viewer changes: take a screenshot with Playwright and check the console is error-free. HTTP 200 proves nothing — a WebGL canvas can be blank.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 55 lines · 987 tokens per session scan A e46316b5f485
codegraph CLAUDE.md is an instructions file published in the GitHub repository RexHung0302/codegraph (0 stars, last pushed 28d ago), licensed MIT. It adds 987 tokens to every session, about $0.0049 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
langchain AGENTS.md
Instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.