The Agent Skills Directory

[SAFE]: The skill provides developer tools (hook-linter.sh, test-hook.sh, validate-hook-schema.sh) designed to assist in the local development and validation of Claude Code plugin hooks. These scripts function as standard CLI utilities for developers.
[SAFE]: Documentation and example scripts (validate-bash.sh, validate-write.sh) actively promote security best practices, teaching users how to detect and block malicious patterns like command injection, path traversal, and secret exposure.
[SAFE]: The use of session-based environment variable persistence via $CLAUDE_ENV_FILE in examples/load-context.sh aligns with the documented and intended architecture of the Claude Code platform.
[SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or persistence mechanisms was found. External integrations mentioned in references/advanced.md (e.g., Slack or database logging) are clearly documented as optional design patterns for advanced users.

hook-development