search-first
Pass
Audited by Gen Agent Trust Hub on Apr 3, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides guidance for installing external packages via package managers like npm or pip when existing solutions are discovered. This is a standard practice in modern software development.
- [DATA_EXFILTRATION]: The skill instructs the agent to check local configuration files at ~/.claude/settings.json and ~/.claude/skills/ to identify existing MCP servers and tools. This is used for environment awareness and to prevent redundant work.
- [PROMPT_INJECTION]: The skill establishes a workflow that involves ingesting data from external registries (npm, PyPI) and web searches, which represents an indirect prompt injection surface. Ingestion points: Web search results, GitHub code snippets, and package registry metadata (SKILL.md). Boundary markers: None explicitly specified for separating research data from agent instructions. Capability inventory: Command execution for package installation and file system access for configuration. Sanitization: No specific validation or escaping mechanisms are described for the gathered research data.
Audit Metadata