The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the GitHub CLI (gh) and runs a local Node.js script (audit-ci-hardening.mjs) to perform security audits and apply repository settings. These operations are within the scope of the skill's stated purpose for CI hardening.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by reading and analyzing workflow and action files from the target repository.
Ingestion points: The script scripts/audit-ci-hardening.mjs reads all files within .github/workflows and .github/actions of the target repository.
Boundary markers: The skill does not implement explicit boundary markers or 'ignore' instructions when reading these files into the agent's context.
Capability inventory: The agent has the capability to write to the local filesystem and modify repository settings via the gh api tool.
Sanitization: The audit script uses regex patterns to identify security gaps and does not execute the content of the files it reads, though the agent's interpretation of the audit results is a potential surface for injection.
[SAFE]: The repository referenced as a blueprint (laststance/skills-desktop) and the tools utilized are associated with the skill's author and represent standard development practices. The recommended security configurations, such as pinning actions to full commit SHAs and using least-privilege permissions, align with industry best practices.

ci-hardening