What is pin-mcp-servers?

A way to record the tools and capabilities exposed by an MCP server in lock files. The record includes tool definitions, the server and protocol, instructions, and whether tools can read, write, delete, use the network, or execute code.

Use it to pin MCP servers, review their capability lists, and verify the pins in CI, the automated build and test process.

It makes unexpected server changes visible in version control and can make a continuous-integration build fail when tools or capabilities change. This helps prevent silently increased access.

pin-mcp-servers — Skill for Claude Code by yotambraun/Toolscore

Install

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.

Any agent

npx skills add yotambraun/Toolscore --skill pin-mcp-servers

Clone the repo

git clone --depth 1 https://github.com/yotambraun/Toolscore

Made for: Claude Code.

Or install toolscore, the plugin that ships this one along with the rest of its 4 skills.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/yotambraun/toolscore/pin-mcp-servers/github.svg)](https://agentmods.dev/skills/yotambraun/toolscore/pin-mcp-servers)

Your own site

<a href="https://agentmods.dev/skills/yotambraun/toolscore/pin-mcp-servers"><img src="https://agentmods.dev/badge/skills/yotambraun/toolscore/pin-mcp-servers/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/yotambraun/toolscore/pin-mcp-servers"><img src="https://agentmods.dev/badge/skills/yotambraun/toolscore/pin-mcp-servers.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 107 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 902 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00107	$0.00902
Opus 5.5	$0.00043	$0.00361
Sonnet 5.5	$0.00021	$0.00180
Haiku 4.5	$0.00011	$0.00090

Measured 5d ago against content hash 407f68a6c2f4, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

pin-mcp-servers 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 5d 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.

plugins/toolscore/skills/pin-mcp-servers/SKILL.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Pin MCP servers

Pinning reads each server's tool list (tools/list) and calls no tools.

0. Toolscore

Use toolscore (1.11 or later; toolscore --version) or uvx tool-scorer in its place. Remote servers need the [http] extra.

1. Which servers

From $ARGUMENTS, or the MCP host config in the repository (.mcp.json, .cursor/mcp.json, .vscode/mcp.json) or the user's client config. List them for the user, with the version each command starts. A command such as npx -y @scope/server without a version can change under the pin at any time; suggest pinning the package version in the command too (npx -y @scope/[email protected]).

2. Pin

One lock file per server keeps diffs readable:

toolscore mcp pin "npx -y @modelcontextprotocol/[email protected]" -o pins/memory.lock.json
toolscore mcp pin --config .mcp.json --server notes -o pins/notes.lock.json

Show the capability manifest it printed (which tools read, write, delete, use the network or execute code) and ask the user to confirm it matches what they expect. If a tool is classified wrongly, write a capability override file and re-pin with --capabilities overrides.json. Commit the lock files.

3. Verify

toolscore mcp verify-pin "npx -y @modelcontextprotocol/[email protected]" --lock pins/memory.lock.json

Exit status 1 means drift. --fail-on capability fails only on a new delete or exec capability; --fail-on never reports without failing. To see exactly what changed between two builds or versions, with text diffs of the descriptions:

toolscore mcp diff "npx -y @modelcontextprotocol/[email protected]" "npx -y @modelcontextprotocol/[email protected]"

4. CI

- uses: yotambraun/toolscore@v1
  with:
    mcp-command: 'npx -y @modelcontextprotocol/[email protected]'
    mcp-pin-file: pins/memory.lock.json
    mcp-pin-fail-on: drift

Or run toolscore mcp verify-pin ... --ci in any job: it writes the report to the GitHub Actions job summary. --sarif drift.sarif (for GitHub code scanning) and --junit drift.xml (for GitLab or Jenkins) write each drift finding, pointing at the lock file. To catch drift before it is committed, add the toolscore-verify-pin pre-commit hook (repo: https://github.com/yotambraun/toolscore, args: ["python server.py", --lock, toolscore.lock.json]). When a change is intended, review the diff with the user, then re-run toolscore mcp pin and commit the new lock file in the same pull request.

Read the full file on GitHub · 76 lines

Changes

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.

5d ago First seen · 76 lines · 107 tokens per session scan A 407f68a6c2f4

Subscribe to this mod's changes

pin-mcp-servers is a skill published in the GitHub repository yotambraun/Toolscore (6 stars, last pushed 6d ago), licensed Apache-2.0. It adds 107 tokens to every session and 902 once invoked, about $0.0004 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 2026-11-04.

CI and CD DevOps and cloud LLM Security ai-agents anthropic function-calling llm-evaluation metrics

pin-mcp-servers next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
pin-mcp-serversyotambraun/Toolscore	Skill	6	6d ago	Apache-2.0	A	107	Claude Code
baby-sitlangchain-ai/open-swe	Skill	11k	today	MIT	A	30	Claude Code, Codex
pr-babysittermblode/agent-skills	Skill	189	4d ago	MIT	A	55	Claude Code
ci-speedupmblode/agent-skills	Skill	189	4d ago	MIT	A	101	Claude Code, Codex
scaffold-climblode/agent-skills	Skill	189	4d ago	MIT	A	52	Claude Code, Codex
audit-harnessalchemiststudiosDOTai/tunacode	Skill	125	1mo ago	MIT	A	69	Claude Code, Codex
devops-expertpersonamanagmentlayer/pcl	Skill	44	14d ago	Apache-2.0	A	47	Claude Code