Skip to main content
Glama

ghe-mcp-gateway

An MCP (Model Context Protocol) server that exposes GitHub Enterprise administration operations as agent tools. Point Claude (Code or Desktop) at it and ask things like "list private repos in the org with no recent pushes" or "who has admin on payments-api?" - the model calls the GitHub API through typed, permissioned tools instead of raw curl.

Built as a working reference for GitHub Enterprise Cloud administration: repository management, access governance, audit-log review, webhook management, and secret-scanning triage.

Why

Admin work is repetitive API calls (access reviews, alert triage, config audits). Wrapping the GitHub REST/GraphQL API as MCP tools lets an agent do the legwork while keeping a human in the loop - and keeps writes behind an explicit safety switch.

Related MCP server: GitHub MCP Server

Tools (21)

Category

Tools

Repository admin

list_repositories, get_repository

Access & teams

list_organization_members, list_teams, list_team_members, list_repo_collaborators, get_repo_permission, list_team_repos, set_repo_collaborator_permission, set_team_repo_permission

Branch protection

get_branch_protection, set_branch_protection*, list_org_rulesets, get_org_ruleset

Security (GHAS)

list_secret_scanning_alerts, list_code_scanning_alerts, list_dependabot_alerts, resolve_secret_scanning_alert*

Audit log

get_org_audit_log

Webhooks

list_org_webhooks

Escape hatch

graphql_query

* write operation - blocked unless GITHUB_MCP_READ_ONLY=false.

Companion skill

.claude/skills/github-admin/SKILL.md packages these tools into admin playbooks - access review/certification, leaked-secret response, repo onboarding & governance, security-posture audit, and access changes. Open Claude Code in this repo and the /github-admin skill is available; it drives the MCP tools with a read-first, least-privilege, confirm-before-write discipline.

Setup

uv sync                       # create venv + install deps
cp .env.example .env          # then add your GITHUB_TOKEN

Suggested token scopes (classic): repo, read:org, admin:org, read:audit_log, admin:org_hook, security_events. For GitHub Enterprise Server, set GITHUB_API_URL to https://<host>/api/v3.

Run

# stdio server (how MCP clients launch it)
uv run github-admin-mcp

# quick manual check with the MCP Inspector
uv run mcp dev src/github_admin_mcp/server.py

Register with Claude Code

claude mcp add github-admin -- uv run --directory /Users/mikeholzinger/src/github_mcp github-admin-mcp

(or add an entry to your client's MCP config pointing at the same command).

60-second demo

# 1. add your token (read-only by default - safe)
cp .env.example .env && $EDITOR .env        # set GITHUB_TOKEN

# 2. register the server + open Claude Code in this repo
claude mcp add github-admin -- uv run --directory "$PWD" github-admin-mcp
claude

# 3. the /github-admin skill is now available. Try, in natural language:
#    "Run an access review on the <org> organization"
#    "Audit the security posture of <org>/<repo>"
#    "Who has admin on <org>/<repo>?"
# The skill calls the MCP tools read-only and reports an auditor-ready summary.

Safety

  • Read-only by default (GITHUB_MCP_READ_ONLY=true); mutating tools refuse until you opt in.

  • No credentials in code - token comes from the environment. .env is git-ignored.

Layout

src/github_admin_mcp/
  client.py   # async REST + GraphQL client (auth, pagination, GHES-aware, read-only guard)
  server.py   # FastMCP server; one @mcp.tool() per operation
docs/
  GITHUB_API_REFERENCE.md   # the endpoint research this server is built on

Example GitHub Actions workflows

Healthcare-oriented admin automation lives in .github/workflows/ with the logic in scripts/ (Python + requests). All follow least-privilege permissions: and pin actions.

Workflow

Trigger

What it does

Access Review (access-review.yml)

Monthly cron + manual

Certifies repo access org-wide; flags outside collaborators, admin grants, write-on-archived; emits a JSON evidence artifact (13-mo retention) and files a tracking issue.

Secret Scanning Alerts (secret-scanning-alerts.yml)

secret_scanning_alert event + hourly sweep

Real-time + safety-net notification to a security Slack channel. Sends metadata only - never the secret value.

Repository Governance (repo-governance.yml)

Org repo-created → repository_dispatch + weekly sweep

Enforces the branch-protection baseline (PR + 2 reviews, Code Owner, signed commits, no force-push/delete, conversation resolution, Dependabot). Dry-run by default; auto-applies to brand-new repos.

Required CI config: secret ORG_ADMIN_TOKEN (org PAT/App), secret SECURITY_SLACK_WEBHOOK, and variable GITHUB_ORG. The scripts honor GITHUB_API_URL for GitHub Enterprise Server.

Roadmap

  • Branch-protection / rulesets tools, team-repo access management

  • Secret-scanning alert resolution, code-scanning + Dependabot alerts

  • Audit-log streaming config; enterprise-level endpoints

  • A companion Claude skill that drives these tools for common admin playbooks

Available Tools

21 tools
get_branch_protectionB

Get the branch-protection configuration for a branch (404 -> unprotected).

Inputs: repo (required), owner (required), branch.

get_org_audit_logA

Query an organization's audit log (requires org-owner token w/ read:audit_log).

phrase: GitHub audit-log search syntax, e.g. 'action:repo.create actor:octocat'. include: web | git | all.

Inputs: org (required), phrase, include. Returns: result

get_org_rulesetA

Get the full definition of a single org ruleset (rules, conditions, bypass actors).

Inputs: org (required), ruleset_id (required).

get_repo_permissionA

Get a user's effective permission level on a repo (admin/write/read/none).

Inputs: repo (required), owner (required), username (required).

get_repositoryC

Get full details for a single repository (settings, visibility, permissions).

Inputs: repo (required), owner (required).

graphql_queryA

Run an arbitrary GraphQL query against GitHub's v4 API.

Useful for nested fetches (e.g. org -> teams -> members) in one round trip.

Inputs: query (required), variables.

list_code_scanning_alertsB

List CodeQL/code-scanning alerts for a repo. state: open | closed | dismissed | fixed.

Inputs: repo (required), owner (required), state. Returns: result

list_dependabot_alertsA

List Dependabot vulnerability alerts. state: open | dismissed | fixed | auto_dismissed.

Inputs: repo (required), owner (required), state. Returns: result

list_organization_membersC

List members of an organization. role: all | admin | member.

Inputs: org (required), role. Returns: result

list_org_rulesetsB

List organization-level rulesets (the modern, scalable branch-protection model).

Inputs: org (required). Returns: result

list_org_webhooksB

List organization-level webhooks (events, active state, target URL).

Inputs: org (required). Returns: result

list_repo_collaboratorsB

List collaborators on a repo with their permission level.

affiliation: outside | direct | all.

Inputs: repo (required), owner (required), affiliation. Returns: result

list_repositoriesA

List repositories for an org or user.

owner: org login or username. owner_type: "org" or "user". type_filter: all | public | private | forks | sources | member (org) / owner|member (user). Returns trimmed repo records (name, visibility, default_branch, archived, pushed_at).

Inputs: owner (required), owner_type, type_filter. Returns: result

list_secret_scanning_alertsA

List secret-scanning alerts for a repo, or for an org if repo is omitted.

state: open | resolved. Requires GHAS/secret-scanning enabled and a token with access.

Inputs: owner (required), repo, state. Returns: result

list_team_membersC

List members of a team. role: all | maintainer | member.

Inputs: org (required), team_slug (required), role. Returns: result

list_team_reposB

List repositories a team can access, with the team's permission on each.

Inputs: org (required), team_slug (required). Returns: result

list_teamsC

List all teams in an organization (slug, name, privacy, parent).

Inputs: org (required). Returns: result

resolve_secret_scanning_alertA

Resolve a secret-scanning alert. WRITE OPERATION.

resolution: false_positive | wont_fix | revoked | used_in_tests. Use after the credential has actually been revoked/rotated. Blocked unless GITHUB_MCP_READ_ONLY=false.

Inputs: repo (required), owner (required), alert_number (required), comment, +1 more.

set_branch_protectionA

Apply a baseline branch-protection policy. WRITE OPERATION.

Requires PRs with required_reviews approvals, optional Code Owner review, blocks force-push/deletion, requires conversation resolution + linear history. Blocked unless GITHUB_MCP_READ_ONLY=false.

Inputs: repo (required), owner (required), branch, enforce_admins, +2 more.

set_repo_collaborator_permissionA

Add or update a collaborator's permission on a repo. WRITE OPERATION.

permission: pull | triage | push | maintain | admin. Blocked unless GITHUB_MCP_READ_ONLY=false.

Inputs: repo (required), owner (required), username (required), permission.

set_team_repo_permissionA

Grant or update a team's permission on a repo. WRITE OPERATION.

permission: pull | triage | push | maintain | admin. Blocked unless GITHUB_MCP_READ_ONLY=false.

Inputs: org (required), repo (required), owner (required), team_slug (required), +1 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.4/5.0

    Scored across 21 tools

    Disambiguation5/5

    Each tool targets a distinct resource or action (e.g., get vs list, different alert types, branch protection vs permissions). There is no ambiguity between tools; they are clearly differentiated by their target and verb.

    Naming Consistency4/5

    Most tools follow a consistent verb_noun pattern (get_*, list_*, set_*, resolve_*) in snake_case. The only outlier is 'graphql_query', which deviates from the pattern but is a generic utility.

    Tool Count4/5

    21 tools is slightly above the typical well-scoped range (3-15), but given the breadth of GitHub admin functionality covered (branch protection, security alerts, permissions, webhooks), it is still reasonable and not excessive.

    Completeness3/5

    The tool set covers reading many resources and some write operations (permissions, branch protection, alert resolution). However, it lacks create/delete operations for repos, teams, and webhooks, which are notable gaps for a full admin gateway.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      A
      maintenance
      An autonomous GitHub management engine that enables control over repositories, branches, security alerts, and Actions workflows through natural language. It utilizes a zero-local-footprint architecture by storing all configuration and audit logs within a private state repository on GitHub.
      1
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Enables asking natural-language questions about any GitHub repository and receiving answers grounded in the actual code, with read-only enforcement and automatic repository fetching.
      1
      7 npm
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables AI assistants to answer natural-language GitHub queries by listing repositories, issues, pull requests, branches, commits, and files, as well as performing writes with dry-run and confirmation safeguards.
      -