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/rutvij26/chess-context/claude-mdgit clone --depth 1 https://github.com/rutvij26/chess-contextWrote 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/rutvij26/chess-context/claude-md)<a href="https://agentmods.dev/instructions/rutvij26/chess-context/claude-md"><img src="https://agentmods.dev/badge/instructions/rutvij26/chess-context/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 | $0.00991 | $0.00991 |
| Opus 5 | $0.00495 | $0.00495 |
| Sonnet 5 | $0.00198 | $0.00198 |
| Haiku 4.5 | $0.00099 | $0.00099 |
Grade A, and why
chess-context 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 — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Status
v0.7.0 — 12 tools live. Opening & training tools added: get_opening_theory, find_opening_gaps, generate_puzzles.
Next: v0.8 (endgame analysis, tilt detection).
Commands
cd mcp-server
npm run build # TS → dist/
npm run dev # tsx (no build)
npm start # dist/index.js
stdio transport — not HTTP. Connect via Claude Desktop.
Structure
mcp-server/src/
index.ts # entry, tool registration
config.ts # env vars
types/index.ts # interfaces + Zod schemas
cache/index.ts # SQLite position cache + LRU player stats cache
cache/sqlite-cache.ts # SQLite-backed eval persistence (survives restarts)
engines/stockfish.ts # WASM UCI wrapper
engines/lichess-eval.ts # cloud eval (tried first)
intelligence/ # pure fns, no I/O
data/ # chesscom-api.ts, lichess-api.ts
tools/ # one handler per tool
Architecture
3 layers: Foundation (engines, data, cache) → Intelligence (pure fns, semantic annotations) → Tools (MCP handlers).
Key Decisions
- Stockfish: WASM via npm, single-threaded, loaded with
createRequirefrom ESM - Init: fire-and-forget after
server.connect()— don't block MCP handshake (~30–60s warmup) - Eval routing: cache → Lichess cloud → Stockfish
- HTTP: axios everywhere (immune to Stockfish's
fetch=nullside effect) - Logs:
console.error()only — stdout is MCP JSON-RPC - TS: strict +
exactOptionalPropertyTypes. Imports use.jsextensions."type": "module"
Env Vars
| Var | Default | Desc |
|---|---|---|
STOCKFISH_DEPTH |
18 |
default depth |
STOCKFISH_QUIET_DEPTH |
10 |
quiet positions (pass 1) |
STOCKFISH_CRITICAL_DEPTH |
16 |
critical positions (pass 2) |
STOCKFISH_TIMEOUT |
30000 |
ms per position |
STOCKFISH_READINESS_TIMEOUT |
90000 |
ms to wait for engine init before rejecting tool calls |
LICHESS_TOKEN |
— | optional, higher rate limits (allows LICHESS_CLOUD_CONCURRENCY=25) |
LICHESS_CLOUD_CONCURRENCY |
10 |
max concurrent cloud eval requests (safe default; raise to 25 with a token) |
EVAL_CACHE_DB |
~/.chess-context/eval-cache.db |
SQLite eval cache path |
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 · 85 lines · 991 tokens per session scan A 9da974e1d8c7
chess-context CLAUDE.md is an instructions file published in the GitHub repository rutvij26/chess-context (4 stars, last pushed 2mo ago), licensed MIT. It adds 991 tokens to every session, about $0.0050 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
mobile CLAUDE.md
Claude Code instructions for lichess-org/mobile, covering claude.md, project overview, development setup, initial setup and install dependencies.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).
llm-context.py CLAUDE.md
Instructions for cyberchitta/llm-context.py, covering claude.md, working notes (gitignored) and draining the field notes.
seekstone CLAUDE.md
Instructions for shaqmughal/seekstone, covering claude.md, what this repo is, commands, the harness itself (run after npm install) and architecture.
Plonk AGENTS.md
AGENTS.md instructions for ostapondo/Plonk, covering agent rules, layout, adding a module, build & verify and code style.
mcp-dockhand copilot-review.instructions.md
Copilot instructions for strausmann/mcp-dockhand: When reviewing pull requests in this repository.