ai-research-agent

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a significant attack surface for Indirect Prompt Injection (Category 8) due to its core loop of reading untrusted external data and using it to generate executable skill files.
  • Ingestion points: The skill explicitly instructs the agent to monitor and scan content from Hacker News, Twitter/X, Reddit, and various technical blogs (documented in the 'Research Sources' section of SKILL.md).
  • Boundary markers: The instructions lack defined delimiters or specific 'ignore embedded instructions' warnings for the data gathered from these external sources.
  • Capability inventory: The agent is empowered to create and modify skill files ('Create/Enhance Skill'), write to the local file system ('Write SKILL.md', 'Add to SKILL_INDEX.json'), and interact with remote repositories ('SKILL.md -> GitHub').
  • Sanitization: There are no requirements for escaping, validation, or filtering of the content retrieved from research sources before it is interpolated into the skill generation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 03:39 PM
Security Audit — agent-trust-hub — ai-research-agent