gitlab-mcp: Instructions file for Codex

AGENTS.md

gitlab-mcp AGENTS.md is an instructions file for Codex, OpenCode from structured-world/gitlab-mcp. It costs 746 tokens per session, scanned A, original, from a forked repository, Apache-2.0.

Repository instructions for GitLab MCP, a tool layer that lets coding agents work with GitLab. They explain supported GitLab versions, API behaviour, compatibility rules, and guidance for reviewers.

In plain words
What is it for?
Use them when changing or reviewing the GitLab MCP repository, especially when adding tool requirements, handling older GitLab versions, or checking fallback behaviour.
Why use it?
They give agents and contributors consistent rules for handling differences between GitLab installations instead of treating the tools as a simple copy of the GitLab API.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md. Also seen: mentions AGENTS.md.

This is structured-world/gitlab-mcp's own configuration. It tells Codex and OpenCode how to work on gitlab-mcp itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything gitlab-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to structured-world/gitlab-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.

Copy the file
curl -O https://raw.githubusercontent.com/structured-world/gitlab-mcp/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/structured-world/gitlab-mcp

Made for: Codex, OpenCode.

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.

agentmods badge for gitlab-mcp AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/structured-world/gitlab-mcp/agents-md/github.svg)](https://agentmods.dev/instructions/structured-world/gitlab-mcp/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/structured-world/gitlab-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/structured-world/gitlab-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.

agentmods 80×15 button for gitlab-mcp AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/structured-world/gitlab-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/structured-world/gitlab-mcp/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 746 This file is loaded in full into every session.
When invoked 746 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin fork From a forked repository.
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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00746 $0.00746
Opus 5.5 $0.00298 $0.00298
Sonnet 5.5 $0.00149 $0.00149
Haiku 4.5 $0.00075 $0.00075

Measured 1mo ago against content hash 22d9b2c48d21, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

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 1mo 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.

AGENTS.md · 55 lines

How it starts

The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Guidance for coding agents and reviewers working on this repository.

GitLab version support

This server supports GitLab 16.0 and later (MIN_SUPPORTED_VERSION in packages/gitlab-mcp/src/services/InstanceCapabilities.ts). It talks to GitLab only through REST API v4 and GraphQL.

Smart MCP, not an API proxy

Tools are an agent-facing layer over GitLab, not a 1:1 proxy: they compose calls, translate parameters, filter and reshape results, and emulate behaviour an older or lower-tier instance lacks. When a capability is missing on some instances, prefer, in order: emulate or degrade (older equivalent, client-side filtering, another endpoint or query, dropping a non-essential field), then report partial effect in the result, and only then hide or refuse, when GitLab itself cannot provide it. Never gate away behaviour a handler already emulates; read the handler before changing its gates.

Rules

  1. Everything is available from 16.0 unless declared otherwise. A tool, action or parameter that needs a newer GitLab declares it in its requirements (default, actions.<action> or parameters.<name>, each { tier, minVersion }). The registry hides whatever the connected instance does not meet.
  2. Declare minVersion only above the floor. A value at or below 16.0 is ignored by the gate and must not be written; it only misleads readers.
  3. Verify every version against GitLab's own sources, never from memory. For each new or changed endpoint, parameter, GraphQL field, argument or fragment type:
    • REST: the {{< history >}} notes in doc/api/*.md, and the Grape optional/requires declaration in lib/api or ee/lib at the release tag (vX.Y.0-ee). Docs often omit the version of a single parameter; the source at the tag is authoritative.
    • GraphQL: doc/api/graphql/reference at the release tag. GitLab rejects the whole query when any selected field, argument or inline-fragment type is unknown, so a single new field gates the entire query and every action using it.
  4. Version-dependent behaviour in handlers uses instanceAtLeast / currentInstance from packages/gitlab-mcp/src/entities/instance-version.ts to pick the native path or the emulation. assertInstanceAtLeast (a clear error) is only for capabilities GitLab cannot provide on that instance, never a substitute for an emulation that is possible.
  5. A 200 response is not proof a setting was applied. GitLab ignores unknown REST parameters and drops license-, add-on- or feature-flag-gated attributes without an error. Where that matters, compare the returned entity with the request (see packages/gitlab-mcp/src/entities/core/duo-settings.ts).

Read the full file on GitHub · 55 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.

  1. 1mo ago First seen · 55 lines · 746 tokens per session scan A 22d9b2c48d21

Subscribe to this mod's changes

gitlab-mcp AGENTS.md is an instructions file published in the GitHub repository structured-world/gitlab-mcp (8 stars, last pushed yesterday), licensed Apache-2.0. It adds 746 tokens to every session, about $0.0030 per session on Opus 5.5. A static security scan graded it A with 0 findings. It comes from a forked repository.

Related

gitlab-mcp AGENTS.md compared with instructions from other repositories

gitlab-mcp AGENTS.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
gitlab-mcp AGENTS.mdstructured-world/gitlab-mcpInstructions file8yesterdayApache-2.0A746Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143ktodayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode