The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs automated Git operations such as git push --force-with-lease, git commit --amend, and git add -f. These actions are necessary for the rebase and CI autofix workflows described in the instructions.
[EXTERNAL_DOWNLOADS]: The skill uses npx to execute the @codevoyant/agent-kit package, which is a tool provided by the skill's author (cloudvoyant). This involves downloading the package from the official NPM registry during execution.
[SAFE]: The skill implements a security check in the commit workflow to warn users if sensitive files like .env or credentials are staged for commitment.
[PROMPT_INJECTION]: The skill contains surfaces for indirect prompt injection in its CI monitoring (references/workflows/ci.md) and rebase conflict resolution (references/workflows/rebase.md) logic.
Ingestion points: Reads CI build logs via gh run view --log-failed and source code file content via cat during conflict resolution.
Boundary markers: None explicitly defined in the prompts to separate external content from instructions.
Capability inventory: Uses git commit, git push --force-with-lease, and automated task runners (format, lint).
Sanitization: No specific sanitization or filtering of log content or code markers is performed before processing.

git