research

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to ingest and process data from untrusted external sources, creating a surface for indirect prompt injection attacks where malicious instructions in external content could influence agent behavior.
  • Ingestion points: Instructions in SKILL.md direct the agent to use gh search on GitHub, browse external URLs (blogs, forums, Reddit, HN, YouTube), and read research papers (arXiv).
  • Boundary markers: The skill lacks explicit boundary markers or instructions to ignore commands or formatting instructions embedded within external documents.
  • Capability inventory: The agent is empowered to modify project files based on research findings ("research inside the loop. Look, apply, re-check"), update pull request descriptions, and spawn parallel subagents to handle scoped research tasks.
  • Sanitization: There are no instructions for sanitizing, escaping, or validating external content before it is interpolated into the agent's context or used to drive code modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 11:01 AM
Security Audit — agent-trust-hub — research