The Agent Skills Directory

[SAFE]: The skill provides operational guidelines for CI/CD interactions without any malicious patterns, obfuscation, or unauthorized data access detected.- [EXTERNAL_DOWNLOADS]: The documentation includes a command to install a dependent skill from the author's official repository using a standard package manager. This is documented neutrally as a routine configuration step.- [COMMAND_EXECUTION]: Instructions authorize the agent to run project-specific formatters and linters to resolve mechanical failures, which is consistent with standard development workflows.- [PROMPT_INJECTION]: The skill processes external data from CI logs and results, creating a surface for indirect prompt injection. However, the potential impact is mitigated by the highly constrained nature of the automated actions. (Ingestion points: CI logs, CI results; Boundary markers: Absent; Capability inventory: git commit, git push, formatter execution, CI retry; Sanitization: Absent).

ci-integration