deep-research

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script (scripts/research.py) to perform multi-step research, check task status, and synthesize reports.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of external dependencies, specifically naming the httpx package via a requirements.txt file.
  • [PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection due to its core functionality of fetching external data.
  • Ingestion points: The agent is designed to search, read, and synthesize information from the open web (SKILL.md).
  • Boundary markers: There are no instructions or delimiters defined to prevent the agent from following malicious instructions that might be embedded in the external websites it researches.
  • Capability inventory: The skill utilizes command execution via python3 to run its research logic (SKILL.md).
  • Sanitization: The instructions do not define any sanitization, filtering, or validation steps for the content retrieved from external sources before it is processed by the model.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 10:51 PM
Security Audit — agent-trust-hub — deep-research