The Agent Skills Directory

[DATA_EXFILTRATION]: The skill reads local files including potential secrets like environment variables and private keys to identify them for cleanup or user warning. No network operations were found that would transmit this data to an external server; all processing occurs locally.
[COMMAND_EXECUTION]: The provided scripts in scripts/clean-repo.mjs and scripts/install-pre-push-hook.sh perform file system operations such as listing, deleting files, and installing git hooks. These actions are intended, documented, and consistent with the skill's primary purpose.
[PROMPT_INJECTION]: The skill ingests untrusted data by reading the content of files within the project directory during the secret-scanning phase in scripts/clean-repo.mjs. This creates an indirect prompt injection surface where malicious content embedded in files could attempt to influence the agent's behavior when the scan results are presented in the agent's context.

cleaner