Borrowing it
Nothing to install: this file belongs to chrischall/opentable-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/chrischall/opentable-mcp/main/CLAUDE.mdgit clone --depth 1 https://github.com/chrischall/opentable-mcpWrote 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/chrischall/opentable-mcp/claude-md)<a href="https://agentmods.dev/instructions/chrischall/opentable-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/chrischall/opentable-mcp/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 5.1 | $0.07084 | $0.07084 |
| Opus 5 | $0.03542 | $0.03542 |
| Sonnet 5 | $0.01417 | $0.01417 |
| Haiku 4.5 | $0.00708 | $0.00708 |
Grade A, and why
opentable-mcp CLAUDE.md scanned grade A with 1 finding 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 2d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Don't add new transport-layer libraries (cycletls, impersonate-curl, Playwright). v0.2 tried those, and each one builds a separate stand-in identity that the rest of the design has to compensate for. The whole point is How it starts
The opening of the file, as written. The whole thing — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — opentable-mcp
Guidance for Claude working in this repo.
TL;DR
OpenTable MCP server with 13 tools (read + write), fronted by a
pluggable browser bridge. Default transport: localhost WebSocket via
@fetchproxy/server — the
companion browser extension is installed separately (Chrome Web Store /
Safari .dmg) rather than embedded in this repo. Opt-in alternative:
OT_BRIDGE=mcp-chrome routes through hangwin/mcp-chrome's HTTP MCP
endpoint instead. Either way, every request rides the user's own browser
session — their cookies, their TLS, their JS context — never ours.
Bridge selection
OT_BRIDGE env var picks the transport:
OT_BRIDGE=websocket(default) — wraps@fetchproxy/server'sFetchproxyServer(WebSocket on 127.0.0.1:37149; override withOT_WS_PORT). The user installs the fetchproxy extension once (Chrome / Safari) instead of loading a per-MCP embedded extension. See https://github.com/chrischall/fetchproxy.OT_BRIDGE=mcp-chrome— talks to hangwin/mcp-chrome athttp://127.0.0.1:12306/mcp(override withOT_MCP_CHROME_URL). Requires mcp-chrome ≥ the release containing PR #348 — thetabUrlparameter onchrome_network_request. Pre-PR mcp-chrome versions are active-tab-only and break credentialed cross-origin fetches. Live-verification of this path is pending the upstream merge.
Commands
npm test— vitest, all mocked, no network. Must stay green.npm run build—tsc --noEmittypecheck + esbuild bundle →dist/bundle.js.npx tsc --noEmit— typecheck only (also runs as part ofnpm run build).node dist/bundle.js— launch the MCP server over stdio (also starts the WS listener).npx tsx scripts/probe-find-slots.ts— live GET round-trip via the extension.npx tsx scripts/probe-favorites-toggle.ts— live add + remove favorite.npx tsx scripts/probe-book-cancel.ts— books and immediately cancels a real reservation. Pick a restaurant that won't mind a 3-second booking.npx tsx scripts/probe-book-cc-cancel.ts— same as above but for a CC-required slot (exercises the preview → book flow).npx tsx scripts/probe-book-cancel-uk.ts— UK-region variant (databaseRegion, country handling).npx tsx scripts/probe-book-cancel-experience.ts— books + cancels a real Experience-mandatory slot (book_preview → book path; targets Cafe Pasqual's).npx tsx scripts/probe-modify-experience.ts— books → modifies (moves time) → cancels a real reservation; the live truth-check for the modify path.npx tsx scripts/probe-find-slots-raw.ts— dumps the raw (unparsed)RestaurantsAvailabilityresponse via thegraphqlcapability — useful for spotting fieldsparse-slots.tscurrently drops. No persisted-query hash involved (see the graphql-capability gotcha in Hot spots/gotchas below).npx tsx scripts/probe-slot-lock.ts [YYYY-MM-DD] [HH:MM]— exercises the slot-lock mutation viabook_preview. Locks a slot for ~90s and makes NO reservation, so it's the safe way to check the write path end-to-end. Prints an explicit diagnosis on 403 (a relay tab without a CSRF token — see gotchas).npx tsx scripts/probe-list-res.ts— dump upcoming reservations; useful after a probe to check for dangling ones.npx tsx scripts/serve-only.ts— raw WS listener that logs every extension frame. Debugging only.npx tsx scripts/e2e-phase-a.ts— read-only smoke (list reservations / profile / favorites).
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.
- 2d ago Changed · +1 lines · +130 tokens per session 90e8f684d457
- 3d ago Changed · +36 lines · +975 tokens per session abb3c8f99127
- 7d ago First seen · 202 lines · 5,979 tokens per session scan A d8c156e25c03
opentable-mcp CLAUDE.md is an instructions file published in the GitHub repository chrischall/opentable-mcp (0 stars, last pushed yesterday), licensed MIT. It adds 7,084 tokens to every session, about $0.0354 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). 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.
vscode buildNext.instructions.md
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
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
langchain AGENTS.md
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.
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
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.