GitHub Actions MCP Server

Manage CI/CD workflows with GitHub Actions through your AI assistant - trigger runs, check status, and debug failures.

April 14, 2026
MCP ServerDevelopment Tools & DevOpsGitHub Actions MCP Server

GitHub Actions MCP Server

Manage GitHub Actions workflows directly from your AI assistant. The GitHub Actions MCP server enables natural language interaction with your CI/CD pipelines, from inspecting workflow runs to debugging failed jobs and managing secrets.

This server is essential for DevOps engineers managing complex multi-repository CI pipelines, developers debugging failing builds, and platform teams maintaining workflow standards across their organization.

What it Does

  • Workflow Management: List and inspect workflows
  • Run Control: Trigger workflow runs and monitor progress
  • Job Debugging: View logs and diagnose failed jobs
  • Artifact Access: Download and inspect build artifacts
  • Secret Management: View and update repository secrets

Installation

# Using npx
npx -y @anthropic-ai/mcp-server-github-actions

# Using Docker
docker run -i --rm -e GITHUB_TOKEN mcp/github-actions

Configuration

{
  "mcpServers": {
    "github-actions": {
      "command": "npx",
      "args": ["-y", "@anthropic-ai/mcp-server-github-actions"],
      "env": {
        "GITHUB_TOKEN": "${GITHUB_TOKEN}"
      }
    }
  }
}

Available Tools

ToolPurposeCategory
list_workflowsList all workflows in a repositoryWorkflows
list_runsList workflow runs with status filteringRuns
get_runGet details of a specific workflow runRuns
trigger_runTrigger a workflow dispatch eventControl
cancel_runCancel a running workflowControl
get_logsFetch logs for a specific jobDebugging
list_artifactsList artifacts from a workflow runArtifacts
list_secretsList repository secrets (names only)Secrets

Common Workflows

Debugging a Failed CI Run

  1. List recent workflow runs and identify the failed one
  2. Get details of the failed run to see which job failed
  3. Retrieve the full logs for the failed job
  4. Ask "Summarize what caused this failure based on the logs"
  5. Fix the issue and re-run the workflow

Automated Deployment Pipeline

  1. Trigger a deployment workflow for a specific environment
  2. Monitor the run progress as it executes
  3. Check each job's status as it completes
  4. Review deployment artifacts after successful completion
  5. Share the run summary with your team

Auditing Workflow Activity

  1. List all workflows across your repositories
  2. Query recent runs for each workflow
  3. Identify workflows with high failure rates
  4. Review logs of recent failures to find patterns
  5. Generate a workflow health report

Example Prompts

  • "Show me the latest workflow runs for main branch"
  • "Why did the CI workflow fail?"
  • "Trigger the deployment workflow for staging"
  • "Show logs from the test job"
  • "List all workflows in this repository"
  • "Find all failed workflows in the last week"

Troubleshooting

IssueLikely CauseSolution
"Workflow not found"Incorrect repository or workflow nameVerify the workflow filename in .github/workflows/
"Token expired"GitHub token needs renewalGenerate a new token with appropriate scopes
"Permission denied"Token lacks required scopesEnsure token has workflow and actions:read scopes
"Workflow disabled"Workflow has been disabled manuallyRe-enable the workflow in GitHub UI or via API

Required Permissions

  • repo - Full repository access
  • workflow - Update GitHub Actions workflows
  • actions:read - View workflow runs and jobs