
from agentclash14
Detect repository CI entrypoints after commits and surface whether checks should be run or automated.
This skill scans a repository after commits to identify CI-related entrypoints (workflows, test scripts, or hooks) and notifies the agent or user when checks should be run, verified, or automated. It acts as a safety and automation primer so teams don't miss build/test steps.
Use this skill as part of post-commit or CI triage flows: when a new commit lands, when reviewing a pull request, or when onboarding a repo to ensure CI checks exist and are executed. It is intentionally conservative — it only surfaces potential CI tasks rather than automatically running destructive steps.
Best used with agent frameworks that handle repository events and can run lightweight checks (Copilot-style assistants, Codex, or internal CI bots).
CI Detector is a skeleton skill with only a brief SKILL.md and no scripts. It claims to detect CI entrypoints after commits and surface whether checks should be run, but provides no implementation, no detection logic, and no actionable instructions. In its current form it offers almost no practical value beyond a placeholder intent.
Pure markdown skill with no security concerns whatsoever. The skill is so minimal it barely qualifies as a skill — it reads more like a design intent or TODO than a usable artifact.