tools-skill-lookup

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] (HIGH): The skill connects to and downloads content from prompts.chat, which is not an authorized or trusted source according to security guidelines. This introduces a supply chain risk.
  • [REMOTE_CODE_EXECUTION] (HIGH): The skill's primary function is to download and save "Helper scripts (Python, shell, etc.)" into the .claude/skills/ directory. This pattern effectively allows an external source to inject executable code into the agent's environment.
  • [COMMAND_EXECUTION] (MEDIUM): Although the skill does not invoke commands directly, it facilitates the persistent storage of scripts that are intended for execution by the agent, creating a persistent attack surface.
  • [INDIRECT PROMPT INJECTION] (LOW): The skill is a major surface for indirect injection.
  • Ingestion points: The get_skill tool retrieves SKILL.md and other documentation from an external API.
  • Boundary markers: Absent. The skill instructions do not specify any delimiters or warnings to ignore instructions within the downloaded content.
  • Capability inventory: File-write operations to the user's home directory (.claude/skills/).
  • Sanitization: Absent. The instructions mandate saving all files exactly as retrieved.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 06:45 PM