research-plus

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it is designed to ingest and process untrusted data from external primary sources.
  • Ingestion points: The agent reads from official documentation, package source code (e.g., node_modules, vendor), and first-party APIs via the Explore and general-purpose tools.
  • Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore instructions that might be embedded in the external content being researched.
  • Capability inventory: The skill allows for writing findings to the local file system (e.g., .scratch/research/) and can trigger other automated actions via slash commands or subagents.
  • Sanitization: There is no evidence of sanitization, validation, or escaping of the external content before it is processed or written to disk.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 01:48 PM
Security Audit — agent-trust-hub — research-plus