skills-master

Warn

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The playwright-pro template includes a script at assets/skill-templates/playwright-pro/scripts/launch-chrome.sh that accesses and clones sensitive browser profile files, including Cookies and Login Data, to maintain session states for analysis.
  • [EXTERNAL_DOWNLOADS]: The update-skills-master template contains logic in assets/skill-templates/update-skills-master/scripts/update_skills_master.py to pull and replace the local skill library using code from the author's remote GitHub repository (github.com/lone-yu-cmd/AI-Coding-Paradigm.git).
  • [COMMAND_EXECUTION]: Multiple components, such as scripts/install.py and assets/skill-templates/add-in-skills-master/scripts/add_skill.py, use the subprocess module to execute shell commands for file operations, package management (npm), and Git operations.
  • [PROMPT_INJECTION]: The skill-creator component (assets/skill-templates/skill-creator/SKILL.md) utilizes strong directive language, such as "CRITICAL: You MUST invoke this skill IMMEDIATELY," to enforce specific AI behavior and path conventions.
  • [COMMAND_EXECUTION]: The context-ai-sync template includes assets/skill-templates/context-ai-sync/scripts/install-hook.sh, which installs a Git pre-commit hook that executes logic automatically before commits.
  • [REMOTE_CODE_EXECUTION]: The update mechanism in update-skills-master allows for the replacement of local executable scripts with code fetched from the vendor's remote repository.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 2, 2026, 11:37 PM