The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the use of the circleci CLI to interact with CircleCI's CI/CD services. This includes reading job outputs, managing workflows, and validating configuration files.
[SAFE]: The skill demonstrates a strong security posture by providing explicit guardrails against credential exfiltration. It instructs the agent to avoid asking for API tokens and to never print raw secret values, ensuring that authentication and data handling remain secure.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it reads potentially untrusted data like job logs and test results from the CircleCI API.
Ingestion points: Job logs, test results, and artifacts are read via the circleci CLI as documented in SKILL.md and references/api-coverage.md.
Boundary markers: Absent.
Capability inventory: Execution of shell-based CLI commands for pipeline management.
Sanitization: Absent; the skill relies on the standard output of the trusted CLI tool for information retrieval.

circleci-cli