Borrowing it
Nothing to install: this file belongs to fetchsandbox/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/fetchsandbox/mcp/main/AGENTS.mdgit clone --depth 1 https://github.com/fetchsandbox/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/fetchsandbox/mcp/agents-md)<a href="https://agentmods.dev/instructions/fetchsandbox/mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/fetchsandbox/mcp/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/fetchsandbox/mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/fetchsandbox/mcp/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00488 | $0.00488 |
| Opus 5.5 | $0.00195 | $0.00195 |
| Sonnet 5.5 | $0.00098 | $0.00098 |
| Haiku 4.5 | $0.00049 | $0.00049 |
Grade A, and why
mcp AGENTS.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 24d 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.
What it actually says
MCP release instructions
For an authorized MCP release, follow the existing distribution procedure in
../docs/mcp-distribution-checklist.md when working in the sandbox repository.
These requirements also apply to the public mirror:
- Build and run the existing MCP tests. Use committed source and exclude secrets and unrelated local changes. Keep package.json, server.json and its nested npm package version aligned. Do not republish an existing npm version.
- Publish npm, verify the served version, and run
scripts/sync-mirror.shfrom that same verified source. Preserve the mirror's registry workflow. - Publish through the existing
publish-registry.ymlworkflow and verify both the official registry version and nested npm version. A successful dispatch alone is not proof of publication. - Smithery is a frozen bundle, not an automatic npm update. For EVERY release,
run
npm run mcpb, inspect the packed manifest version, authentication text, tool input schemas and annotations, then start the unpacked bundle and call MCP initialize/tools/list. Confirm its JavaScript matches the npm artifact. Publish withsmithery mcp publish ./fetchsandbox-mcp.mcpb -n fetchsandbox/mcp. Verify SUCCESS, the new deployment/bundle ID, and the actual public tool list. Derive expected tools from the running bundle; do not freeze a tool count. Local and hosted transports intentionally have different advertised tool sets. - Glama: Sync Server under Repository, verify the fetched commit, then Build & Release under Dockerfile. Verify the actual test result and released version. Record external builder failures as blocked, never as a passed release.
- Check other directories individually. Preserve existing submission workflows; do not infer freshness from npm or the official registry alone.
Keep API keys optional for startup/discovery, while describing authentication requirements for application planning, verification and managed local tunnels. Never upload real keys or dummy keys pretending to authenticate. Preserve the existing bundle builder and its inputSchema/annotation compatibility handling.
Cached API responses may lag the public page. Check the page and a fresh read before concluding that publishing failed; never blindly republish a duplicate. Save per-surface version, revision/deployment ID, URL, test outcome and blocker. The final update must distinguish verified, pending and blocked surfaces.
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.
- 24d ago First seen · 38 lines · 488 tokens per session scan A 70766d5a5d8d
mcp AGENTS.md is an instructions file published in the GitHub repository fetchsandbox/mcp (1 stars, last pushed 18d ago), licensed MIT. It adds 488 tokens to every session, about $0.0020 per session on Opus 5.5. 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 2026-10-13.
mcp AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| mcp AGENTS.mdfetchsandbox/mcp | Instructions file | 1 | 18d ago | MIT | A | 488 | Codex, OpenCode |
| skill-doctor git-workflow.instructions.mdevilstar2016/skill-doctor | Instructions file | 16 | 8d ago | MIT | A | 75 | GitHub Copilot |
| skill-doctor github-automation.instructions.mdevilstar2016/skill-doctor | Instructions file | 16 | 8d ago | MIT | A | 72 | GitHub Copilot |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | yesterday | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | yesterday | MIT | A | 5,516 | GitHub Copilot |