The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git and node to perform local repository analysis. This is essential for identifying sensitive data in diffs and files.
[PERSISTENCE_MECHANISMS]: The install-git-hook.mjs script installs a git pre-commit hook in the repository's .git/hooks directory. This is a documented persistence mechanism used to automate the execution of security scans during the development workflow.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository. Ingestion points: The scan-sensitive.mjs script reads file content and git diffs to identify secrets. Boundary markers: No specific delimiters or "ignore instructions" warnings are used around the scanned content in the output. Capability inventory: The skill maintains the ability to execute git and node commands to perform its analysis. Sanitization: Matches for credentials, tokens, and private keys are masked in the output to prevent accidental secondary exposure.

public-repo-guard