GitHub Actions MCP Server
Manage CI/CD workflows with GitHub Actions through your AI assistant - trigger runs, check status, and debug failures.
GitHub Actions MCP Server
This page documents GitHub's official MCP server, github/github-mcp-server, configured with its actions toolset. The server is distributed as the container image ghcr.io/github/github-mcp-server and as the remote endpoint https://api.githubcopilot.com/mcp/.
The actions toolset is not part of the default toolsets, so it must be enabled explicitly.
What it Does
- Workflow and run inspection: List workflows, workflow runs, jobs, and artifacts
- Run details: Get details for a workflow, run, job, or artifact
- Run control: Trigger workflow runs and perform run actions
- Job debugging: Retrieve logs for a job or for all failed jobs in a run
Installation
# Docker (stdio) with the actions toolset
docker run -i --rm \
-e GITHUB_PERSONAL_ACCESS_TOKEN \
-e GITHUB_TOOLSETS=actions \
ghcr.io/github/github-mcp-server
The remote server is available at https://api.githubcopilot.com/mcp/ for hosts that support remote MCP servers.
Configuration
{
"mcpServers": {
"github-actions": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"GITHUB_PERSONAL_ACCESS_TOKEN",
"-e",
"GITHUB_TOOLSETS=actions",
"ghcr.io/github/github-mcp-server"
],
"env": {
"GITHUB_PERSONAL_ACCESS_TOKEN": "YOUR_TOKEN"
}
}
}
}
GITHUB_TOOLSETS selects the toolsets to enable. Set it to actions to expose only the GitHub Actions tools; combine with --tools for individual tools.
Available Tools
| Tool | Purpose | Category |
|---|---|---|
actions_list | List workflows, workflow runs, jobs, or artifacts in a repository | Workflows |
actions_get | Get details of a workflow, workflow run, job, or artifact | Runs |
actions_run_trigger | Trigger a workflow run and perform run actions | Control |
get_job_logs | Get logs for a workflow job, or all failed jobs in a run | Debugging |
The full tool list and parameters are in the upstream README.
Common Workflows
Debugging a Failed CI Run
- Use
actions_listto find the failed run - Use
actions_getto inspect the run - Use
actions_listto find the failed job - Use
get_job_logsto retrieve the failed job logs - Ask "Summarize what caused this failure based on the logs"
Triggering a Deployment Workflow
- Use
actions_listto confirm the workflow ID or file name - Use
actions_run_triggerto run the workflow on the target ref - Pass workflow inputs when the workflow accepts them
- Use
actions_listto monitor the run - Use
actions_getto inspect the run and its jobs
Auditing Workflow Activity
- Use
actions_listto list the repository's workflows - Use
actions_listto query recent runs - Use
actions_getto inspect runs of interest - Use
get_job_logson failed runs to find patterns - Generate a workflow health report
Example Prompts
- "Show me the latest workflow runs for the 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
| Issue | Likely Cause | Solution |
|---|---|---|
| "Workflow not found" | Incorrect owner, repository, or workflow ID | Verify the owner, repository, and workflow ID or file name (for example ci.yaml) |
| "Bad credentials" | Missing or expired token | Set GITHUB_PERSONAL_ACCESS_TOKEN to a valid token |
| "Permission denied" | Token lacks the required scope | The Actions tools require the repo OAuth scope; grant equivalent repository access |
| No Actions tools are available | GITHUB_TOOLSETS does not include actions | Set GITHUB_TOOLSETS=actions; the actions toolset is not enabled by default |
| Server fails at startup | An invalid tool name was passed | Tool names must match exactly, for example get_job_logs |
Required Permissions
repo- OAuth challenge scope used by theactions_list,actions_get,actions_run_trigger, andget_job_logstools- Repository access to the workflows, runs, jobs, and artifacts being queried
Sources
Related Articles & Guides
Accessibility Scanner MCP Server
Accessibility Scanner MCP servers enable AI models to perform WCAG compliance checks, capture annotated screenshots, and generate detailed accessibility reports.
Auth0 MCP Server
Auth0 MCP server enables AI models to interact with Auth0 identity and access management, including applications, APIs, actions, logs, and forms.
Azure DevOps MCP Server
The Azure DevOps MCP Server enables AI models to interact with Azure DevOps, providing capabilities for managing work items, pull requests, pipelines, and more.