The Agent Skills Directory

[SAFE]: The skill provides documentation for interacting with the official Anthropic claude CLI tool for legitimate coding tasks such as refactoring, debugging, and code review.
[COMMAND_EXECUTION]: The skill documents command-line usage of the claude tool. The execution patterns described are consistent with the intended purpose of facilitating terminal-based coding assistance.
[DATA_EXFILTRATION]: The skill mentions the use of sensitive context (codebases, git diffs) and the ANTHROPIC_API_KEY, but only within the scope of interacting with the official service. No unauthorized exfiltration patterns to third-party domains were found.

claude-code