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/vinihcrosa/gitlab-mcp/agents-mdgit clone --depth 1 https://github.com/vinihcrosa/gitlab-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/vinihcrosa/gitlab-mcp/agents-md)<a href="https://agentmods.dev/instructions/vinihcrosa/gitlab-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/vinihcrosa/gitlab-mcp/agents-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 | $0.05411 | $0.05411 |
| Opus 5 | $0.02705 | $0.02705 |
| Sonnet 5 | $0.01082 | $0.01082 |
| Haiku 4.5 | $0.00541 | $0.00541 |
Grade A, and why
gitlab-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 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.
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 — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Guia de Orientação para Agentes de Código
Este documento existe para orientar qualquer agente de código (humano ou LLM) que precise trabalhar neste repositório. Ele descreve o que o produto faz, como o repositório está organizado, quais comandos existem de verdade, quais invariantes de domínio o código-fonte estabelece, e quais ações um agente não deve executar. Leia este arquivo inteiro antes de fazer qualquer alteração, pois várias regras aqui descritas são invariantes rígidos derivados diretamente do código, e violá-las quebra o servidor de formas silenciosas e difíceis de depurar.
1. O que este produto faz
O gitlab-mcp é um servidor MCP (Model Context Protocol) que roda sobre stdio e funciona como um proxy fino sobre a REST API v4 de uma instância GitLab CE self-hosted. O escopo é deliberadamente pequeno (é um MVP com um objetivo só): navegar merge requests e deixar review inline sem abrir o browser. Ler o estado da CI entra nesse objetivo — a medição de uso real (157 de 632 chamadas gh) mostrou que o fluxo trava exatamente quando a pipeline quebra; ver docs/adr/2026-08-12-pipelines-re-enter-scope-on-measured-usage.md. Escrever em CI não entra: nada dispara, cancela ou re-roda pipeline, e nada lê variável ou secret. Fora de escopo também — issues, criação/merge de MR, aprovações, ou recursos Premium/Ultimate.
O servidor expõe exatamente 13 tools MCP, registradas em ordem em src/tools/index.ts:
whoami— identidade do token (src/tools/whoami.ts).list_my_projects— projetos onde o usuário é membro (src/tools/projects.ts).list_my_authored_mrs— MRs criados pelo usuário (src/tools/mrs.ts).list_mrs_awaiting_my_review— MRs abertos onde o usuário é reviewer (src/tools/mrs.ts).get_mr— detalhe completo de um MR (src/tools/mrs.ts).get_mr_diff— diff em texto com numeração explícitaold=/new=(src/tools/diff.ts).list_mr_discussions— threads de comentário de um MR (src/tools/discussions.ts).comment_on_mr— comentário geral no MR (escrita,src/tools/write.ts).comment_on_mr_line— thread de review ancorada em linha do diff (escrita,src/tools/write.ts).reply_to_mr_discussion— resposta em thread existente (escrita,src/tools/write.ts).get_mr_pipeline— pipeline mais recente do MR e seus jobs (leitura,src/tools/pipelines.ts).get_job_log— log do job, limpo e cortado pela cauda (leitura,src/tools/pipelines.ts).list_pipelines— pipelines do projeto, filtradas na API (leitura,src/tools/pipelines.ts).
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 First seen · 188 lines · 5,411 tokens per session scan A 7674f04d64e3
gitlab-mcp AGENTS.md is an instructions file published in the GitHub repository vinihcrosa/gitlab-mcp (0 stars, last pushed 3d ago), licensed MIT. It adds 5,411 tokens to every session, about $0.0271 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
gitlab-mcp CLAUDE.md
Instructions for ttpears/gitlab-mcp, covering claude.md, commands, architecture, project conventions (from .cursorrules) and release flow.
gitlab-mcp copilot-instructions.md
Instructions for structured-world/gitlab-mcp, covering github copilot instructions for gitlab-mcp, project overview, commit message rules (critical), required commit types and commit message guidelines.
llm-context.py CLAUDE.md
Instructions for cyberchitta/llm-context.py, covering claude.md, working notes (gitignored) and draining the field notes.
openrouter-mcp-multimodal 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).
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
Instructions for ostapondo/Plonk, covering agent rules, layout, adding a module, build & verify and code style.