codex-deep-research

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of instructional markdown and configuration files with no executable scripts or code. It defines a robust research methodology focused on verifying source quality and synthesizing information safely.- [SAFE]: To address the risks inherent in processing untrusted data from the web, the skill implements a strict output contract. It requires the use of structured templates (assets/report-template.md) and clear evidence boundaries (references/citation-rules.md) to distinguish between facts, inferences, and recommendations. This structural isolation effectively mitigates potential indirect prompt injection concerns.- [SAFE]: No instances of obfuscation, hardcoded credentials, unauthorized network access, or persistence mechanisms were found. The skill operates within expected research task boundaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 09:33 AM