skills/thaitype/chief/chief-install/Gen Agent Trust Hub

chief-install

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: Clones the 'chief' framework from the author's official GitHub repository and executes an installation script (setup.sh) located in the temporary folder.
  • [EXTERNAL_DOWNLOADS]: Fetches framework templates and scripts from the author's official GitHub repository (thaitype/chief).
  • [COMMAND_EXECUTION]: Utilizes system commands such as git, bash, cp, ln, and rm to manage installation files and create integrations for development tools like Claude Code and GitHub Copilot.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts user-provided arguments for the framework version and configuration values for agent models, which are interpolated into commands and project files.
  • Ingestion points: User-provided version argument and interactive prompts for agent type, installation mode, and model names (SKILL.md).
  • Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are applied to the processed inputs (SKILL.md).
  • Capability inventory: The skill performs repository cloning, shell script execution, and multiple file system modifications including write, copy, and symbolic link operations (SKILL.md).
  • Sanitization: Input validation or escaping for the version string and model names is not explicitly defined in the instructions (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 11:41 AM