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/vugarfamiloglu/graph-x/claude-mdgit clone --depth 1 https://github.com/vugarfamiloglu/graph-xWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/vugarfamiloglu/graph-x/claude-md)<a href="https://agentmods.dev/instructions/vugarfamiloglu/graph-x/claude-md"><img src="https://agentmods.dev/badge/instructions/vugarfamiloglu/graph-x/claude-md.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.00887 | $0.00887 |
| Opus 5 | $0.00443 | $0.00443 |
| Sonnet 5 | $0.00177 | $0.00177 |
| Haiku 4.5 | $0.00089 | $0.00089 |
Grade A, and why
graph-x 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 4d ago.
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 — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Graph X
Local knowledge graph tool for any project (formerly "bilik-qrafi"): incremental indexer + SQLite graph DB + MCP server + interactive graph UI. Zero npm dependencies (uses portable Node 24 in .tools/node/, built-in node:sqlite). Goal: cut Claude Code token usage by querying the graph instead of reading files.
Architecture
- src/config.js — ignore rules, file classification, settings loader (
graphx.config.json, see docs/CONFIGURATION.md) - src/scan.js — recursive walker, .gitignore support
- src/parsers.js — regex-based import/export/link extraction (JS/TS, Python, Markdown, CSS, HTML, package.json) + import resolution
- src/extractors.js — zero-dep PDF/DOCX text extraction (node:zlib; FlateDecode + ToUnicode CMaps, ZIP central directory)
- src/tfidf.js — tokenizer, TF-IDF vectors, cosine similarity, semantic edge discovery
- src/summary.js — structural summaries (Azerbaijani)
- src/indexer.js — orchestrates scan → parse → store; hash-based incremental; rebuilds edges + semantic pairs
- src/db.js — SQLite schema: files / edges / meta (
.graph/graph.dbin target project) - src/graph.js — query layer: project map, search, connections, impact BFS, graph JSON
- src/mcp-server.js — hand-rolled newline-JSON-RPC stdio MCP server (7 tools)
- src/server.js — HTTP UI server + SSE live updates + fs.watch auto-reindex (port 7341)
- src/setup.js — attach the tool to another project (.mcp.json + CLAUDE.md + launch.json + first index; migrates legacy bilik-qrafi/galaxy-x entries)
- src/snapshot.js — static SVG renderer (headless force layout, top-N cap for big graphs)
- src/github.js — GitHub repo download over HTTPS (no git; zip extractor from extractors.js)
- ui/ — canvas force-directed graph interface (vanilla JS, no libraries), dual-theme neumorphic chrome
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.
- 4d ago First seen · 41 lines · 887 tokens per session scan A ed8e213a6f2b
graph-x CLAUDE.md is an instructions file published in the GitHub repository vugarfamiloglu/graph-x (0 stars, last pushed 1mo ago), licensed MIT. It adds 887 tokens to every session, about $0.0044 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
amfs CLAUDE.md
Instructions for raia-live/amfs, covering amfs memory — agent instructions, available mcp tools, identity, brain tools (agent-scoped) and shared knowledge tools.
ScalaSemantic AGENTS.md
AGENTS.md instructions for MercurieVV/ScalaSemantic, covering agents.md instructions and quick start.
grok-faf-mcp CLAUDE.md
Claude Code instructions for Wolfe-Jam/grok-faf-mcp, covering claude.md — grok-faf-mcp, what this is, stack and context.
sonde AGENTS.md
AGENTS.md instructions for anishmoncivarghese/sonde, covering project history, recording decisions, handing off active work, sonde and picking up this project (any agent, any vendor).
clawstash CLAUDE.md
Claude Code instructions for fo0/clawstash, covering claude.md -- project guide, session start -- read order, workflow triggers, output languages and performance / modes.
TechDebtMCP copilot-instructions.md
Instructions for PierreJanineh/TechDebtMCP, covering ai agent instructions for tech debt mcp, project overview, architecture, copilot pr review workflow and before creating a pr.