Feature Request: Add tool to fetch job logs by job name (avoid listing all jobs)

[cisco-krsna]

Describe the feature or problem you’d like to solve

Currently, the GitHub MCP Server exposes the following tools for working with workflow runs and jobs:

• mcp0_list_workflow_runs
• mcp0_list_workflow_jobs
• mcp0_get_job_logs

To fetch logs for a particular job, an agent must:

1. Call mcp0_list_workflow_runs to find the relevant run.
2. Call mcp0_list_workflow_jobs with the run_id.
3. Iterate through hundreds or even thousands of jobs to locate a job by its name.
4. Call mcp0_get_job_logs with the job_id.

This approach works for small workflows but becomes inefficient and impractical for large workflows (e.g. 1,000+ jobs). Agents often struggle to handle such large responses, and passing the entire job list into context is wasteful.

---

Proposed solution

Introduce a new tool such as:

• mcp0_get_job_logs_by_name
  • Inputs: run_id, job_name
  • Behavior: Server internally paginates jobs for the given run, finds the job with the given name, and returns its logs.

Alternatively:

• mcp0_get_job_by_name
  • Inputs: run_id, job_name
  • Behavior: Server returns the job_id only, which can then be passed to the existing mcp0_get_job_logs.

Benefits:

• Efficient for large workflows with hundreds/thousands of jobs.
• Reduces token/context overhead for AI agents.
• Keeps API usage minimal (avoids unnecessary job list retrieval).
• Provides a clearer and more direct developer experience.

---

Example prompts or workflows (for tools/toolsets only)

1. Fetch logs from a specific job in the latest successful run

   {
     "tool": "mcp0_get_job_logs_by_name",
     "arguments": {
       "run_id": 678901234,
       "job_name": "integration-tests"
     }
   }
   

2. Get job ID by name

{
 "tool": "mcp0_get_job_by_name",
 "arguments": {
   "run_id": 678901234,
   "job_name": "build-backend"
 }
}

Simplify common workflow automation

• “Fetch the logs for the lint job in the most recent successful run of ci.yml”
• “Retrieve job IDs for all deploy-* jobs in a workflow run.”

---

Additional context

• Large enterprise workflows often include hundreds or thousands of jobs. Without server-side filtering by job name, AI agents or clients must inefficiently fetch and parse the entire list.
• Adding a server-side helper tool would significantly improve scalability and usability.

[kaihendry]

Current actions X-MCP-Toolsets MCP is really lacking, this post has some good suggestions.

My ask is to succinctly expose workflow Inputs in the response.

[github-actions]

This issue is stale because it has been open for 30 days with no activity. Leave a comment to avoid closing this issue in 60 days.

[github-actions]

This issue was closed because it has been inactive for 60 days since being marked as stale.