Security audit
Security checks for vulnerabilities and agentic risk
This skill is a documentation helper whose broad activation wording may cause extra doc edits, but its behavior is disclosed and limited to project documentation updates.
Install this if you want the agent to proactively keep docs aligned with user-facing code changes. Review proposed documentation edits for relevance, especially on internal refactors or test-only changes where documentation may not actually need to change.
The skill metadata uses broad trigger language such as 'automatically update' and lists many common development events, which can cause over-invocation in ordinary coding workflows. In an agentic environment, ambiguous activation boundaries can lead to unintended documentation edits, unnecessary file modifications, and noisy commits that affect repository integrity and developer trust.
The body reinforces autonomous detection and execution without defining when the skill should not activate, making it prone to matching many routine code changes. This increases the chance of unauthorized or low-signal documentation churn, and in multi-skill systems could cascade into broader automated actions based on incorrect assumptions about required doc updates.
No suspicious patterns detected.
