// SCAN REPORT

GitHub MCP Server

87 tools. 35 can modify or destroy data without limits.

7 destructive tools with no built-in limits. Policy required.

Last updated: 3 Jul 2026

35 can modify or destroy data

52 read-only

87 tools total

Community server · catalogue entry verified 03/07/2026

How to control GitHub MCP Server ↓

TOOL MAP

What GitHub MCP Server exposes to your agents

Read (52) Write / Execute (28) Destructive / Financial (7)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous GitHub MCP Server tools

Destructive · High cancel_workflow_run Cancelling a workflow run is an irreversible action: a cancelled run cannot be resumed and any in-progress work is lost. Destructive · High delete_branch Deleting a branch removes version control history and references irreversibly. Destructive · High delete_branch_protection This tool removes branch protection rules, which are security controls that prevent direct pushes, enforce code reviews, and require status checks. Destructive · High delete_file delete_file removes data that cannot be recovered without access to version history or backups.

35 of GitHub MCP Server's 87 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control GitHub MCP Server

PolicyLayer is an MCP gateway — it sits between your AI agents and GitHub MCP Server, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "cancel_workflow_run": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "add_collaborator": {
    "limits": [
      {
        "counter": "add_collaborator_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "get_authenticated_user": {
    "limits": [
      {
        "counter": "get_authenticated_user_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register GitHub MCP Server — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON GITHUB →

Instant setup, no code required.

FULL CATALOGUE

All 87 GitHub MCP Server tools

DESTRUCTIVE 7 tools

Destructive cancel_workflow_run cancel_workflow_run Destructive delete_branch delete_branch Destructive delete_branch_protection delete_branch_protection Destructive delete_file delete_file Destructive delete_label delete_label Destructive delete_webhook delete_webhook Destructive remove_collaborator remove_collaborator

EXECUTE 1 tools

Execute trigger_workflow trigger_workflow

WRITE 27 tools

Write add_collaborator add_collaborator Write create_branch create_branch Write create_commit_status create_commit_status Write create_gist create_gist Write create_issue create_issue Write create_issue_comment create_issue_comment Write create_label create_label Write create_or_update_file create_or_update_file Write create_pr_review create_pr_review Write create_pull_request create_pull_request Write create_release create_release Write create_repository create_repository Write create_review_comment create_review_comment Write create_webhook create_webhook Write fork_repository fork_repository Write mark_notifications_read Mark all notifications as read. Write mark_thread_read Mark a specific notification thread as read. Write merge_pull_request merge_pull_request Write push_files push_files Write reply_to_review_comment reply_to_review_comment Write star_repository star_repository Write unstar_repository unstar_repository Write update_branch_protection update_branch_protection Write update_gist update_gist Write update_issue update_issue Write update_pull_request update_pull_request Write update_pull_request_branch update_pull_request_branch

READ 52 tools

Read get_authenticated_user Get information about the currently authenticated user. Read get_branch_protection get_branch_protection Read get_check_run get_check_run Read get_commit get_commit Read get_discussion get_discussion Read get_discussion_comments get_discussion_comments Read get_file_contents get_file_contents Read get_gist Get a specific gist with its file contents. Read get_issue get_issue Read get_latest_release get_latest_release Read get_organization Get detailed information about an organization. Read get_project_v2 get_project_v2 Read get_pull_request get_pull_request Read get_release_by_tag get_release_by_tag Read get_repository get_repository Read get_repository_tree get_repository_tree Read get_tag get_tag Read get_user Get public information about a GitHub user. Read list_branches list_branches Read list_check_runs list_check_runs Read list_collaborators list_collaborators Read list_commit_statuses list_commit_statuses Read list_commits list_commits Read list_discussion_categories list_discussion_categories Read list_discussions list_discussions Read list_gists list_gists Read list_issue_comments list_issue_comments Read list_issues list_issues Read list_labels list_labels Read list_notifications list_notifications Read list_org_members list_org_members Read list_org_projects list_org_projects Read list_pr_commits list_pr_commits Read list_pr_files list_pr_files Read list_pr_reviews list_pr_reviews Read list_project_v2_items list_project_v2_items Read list_projects_v2 list_projects_v2 Read list_pull_requests list_pull_requests Read list_releases list_releases Read list_repo_projects list_repo_projects Read list_repositories list_repositories Read list_review_comments list_review_comments Read list_starred_repos list_starred_repos Read list_user_orgs list_user_orgs Read list_webhooks list_webhooks Read list_workflow_runs list_workflow_runs Read list_workflows list_workflows Read search_code search_code Read search_issues search_issues Read search_repositories search_repositories Read search_users search_users

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about GitHub MCP Server

Can an AI agent delete data through the GitHub MCP Server MCP server? +

Yes. The GitHub MCP Server server exposes 7 destructive tools including cancel_workflow_run, delete_branch, delete_branch_protection. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through GitHub MCP Server? +

The GitHub MCP Server server has 27 write tools including add_collaborator, create_branch, create_commit_status. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach GitHub MCP Server.

How many tools does the GitHub MCP Server MCP server expose? +

87 tools across 4 categories: Destructive, Execute, Read, Write. 52 are read-only. 35 can modify, create, or delete data.

How do I enforce a policy on GitHub MCP Server? +

Register the GitHub MCP Server MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every GitHub MCP Server tool call.

Deterministic rules across all 87 GitHub MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON GITHUB →

Instant setup, no code required.

87 GitHub MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.