The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs subprocess.run across several components like run.py and scan.py to coordinate local scripts and execute the official gh CLI. These operations are scoped to the skill's local directory and verified external tools.
[EXTERNAL_DOWNLOADS]: The data pass fetches CI run history and job logs from GitHub via the gh tool. These downloads are performed over a trusted service domain and are essential for the tool's diagnostic and measurement functions.
[PROMPT_INJECTION]: The skill anticipates indirect prompt injection from third-party CI logs and mitigates the risk by providing clear instructional boundaries to the agent in SKILL.md and by implementing sanitization functions such as _fence_safe to neutralize potentially malicious formatting in generated reports.

ci-speedup