The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes standard system commands (ls, cat) to discover project structure and read configuration files related to CI/CD platforms such as GitHub Actions, GitLab CI, CircleCI, Jenkins, and Docker.
[DATA_EXFILTRATION]: As part of its auditing and risk assessment function, the skill identifies and documents where secrets and sensitive environment variables are referenced within pipeline configurations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources such as .github/workflows/*.yml, Dockerfile, and other CI/CD manifests. This represents a potential attack surface where instructions embedded in these files could influence agent behavior.
Ingestion points: Pipeline and deployment configuration files read via cat in Step 1.
Boundary markers: Absent.
Capability inventory: Bash, Read, WebFetch, WebSearch, AskUserQuestion.
Sanitization: No sanitization or validation of the ingested file content is performed.

relay-recon