wiki-researcher

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • Command Execution (LOW): The skill requires the agent to execute shell commands (git remote get-url origin and git rev-parse --abbrev-ref HEAD) to determine the repository state. While common for development tools, this grants the skill local command execution capabilities.\n- Indirect Prompt Injection (LOW): The skill is designed to ingest and interpret code from arbitrary repositories, creating a surface for indirect prompt injection.\n
  • Ingestion points: Implementation files and documentation within the user's codebase.\n
  • Boundary markers: Absent. The skill provides no instructions to ignore or treat comments/data within the analyzed files as untrusted.\n
  • Capability inventory: Extensive file reading and local git command execution.\n
  • Sanitization: Absent. The skill iterates through code paths and summaries without escaping or validating the content of the analyzed files.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 09:38 PM