Summary

The get_job_logs tool currently returns complete job logs, which can be extremely long and inefficient when used as MCP tool context for LLMs. Adding log truncation options like tail_lines parameters would help focus on relevant failure information.

Problem

When debugging CI/CD failures using the get_job_logs tool in LLM conversations:

1. Context Overflow: Complete job logs can easily exceed LLM context limits
2. Information Overload: Most debugging scenarios only need the final portion where failures occur
3. Poor User Experience: Long logs make it harder to identify specific failure points

Proposed Solution

Add optional truncation parameters to the get_job_logs tool:

• tail_lines: Return last N lines (most useful for failure debugging)

Example usage:

{
  "name": "get_job_logs",
  "arguments": {
    "owner": "example",
    "repo": "repo", 
    "job_id": 12345,
    "tail_lines": 100,
    "return_content": true
  }
}

This enhancement aligns with existing concerns about LLM context efficiency, similar to Issue #142 regarding excessive context size.