The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill uses the gh pr merge --admin command to bypass branch protection rules and CI status checks. While the skill includes a checkpoint to ask the user before performing this action, this capability allow for the circumvention of established repository security policies.
[INDIRECT_PROMPT_INJECTION]: The scripts/audit-repo.sh script fetches and decodes the Base64 content of README files from target repositories (via gh api). This content is then processed and summarized for the agent, representing an ingestion point for untrusted data that could contain malicious instructions designed to influence the agent's subsequent remediation actions. * Ingestion points: scripts/audit-repo.sh line 118 (README content decoding). * Boundary markers: None identified in the script output for the agent. * Capability inventory: gh api (write access), git push, gh pr create, gh pr merge --admin. * Sanitization: None; the agent processes the resulting audit reports and the repository content directly during the fix phase.
[EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to fetch governance templates and configuration files from cislunarspace/CODE-core, which is an external repository not associated with the skill author.
[COMMAND_EXECUTION]: The skill performs extensive shell command execution, including gh api calls that can modify repository settings, rulesets, and protection configurations. It also performs standard Git operations (clone, checkout, push) and GitHub CLI operations (pr create, pr merge).

github-governance