The Agent Skills Directory

[COMMAND_EXECUTION]: The script executes npx @tiangong-ai/cli@0.0.19, which is a tool provided by the skill's author (tiangong-ai). This is standard behavior for a CLI-based skill.\n- [EXTERNAL_DOWNLOADS]: The skill uses npx to fetch the @tiangong-ai/cli package. Since the package belongs to the skill's own vendor, this is considered a safe and intended dependency resolution.\n- [CREDENTIALS_UNSAFE]: The skill handles authentication tokens (bearer tokens and API keys) by forwarding them as flags to the CLI tool. It includes a dry_run mode that explicitly masks credentials, following security best practices for debug outputs. Environment variables are loaded from .env files if specified, which is a standard method for secret management.\n- [DATA_EXPOSURE]: The skill restricts searches specifically to the course source, preventing unintended access to other knowledge-base categories like textbook or edu. It processes user-provided JSON but uses jq for safe parsing and validation.

tiangong-kb-course-search