deep-research — security audits

Review recommended

Independent security partners scan this skill's files for malicious content, prompt injection and risky patterns. Verdicts below come from the skills.sh audit program; each partner links to its full report.

• Gen Agent Trust Hub

  pass

  The 'deep-research' skill is a sophisticated research orchestration framework that uses a lead agent and subagents to perform multi-pass research, evidence tracking, and counter-review. It includes a robust 'Source Accessibility Policy' designed to prevent the misuse of user private data for 'circular verification' while allowing the use of user-provided exclusive sources for third-party research. The skill's architecture naturally mitigates indirect prompt injection risks by filtering web data through subagent-generated notes before the lead agent processes it.

  Risk: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTIONMay 29, 2026Full report

• Socket

  pass

  No alerts

  May 29, 2026Full report

• Snyk

  warn

  Risk: MEDIUM · 2 issues

  Risk: MEDIUMMay 29, 2026Full report

• Runlayer

  warn

  7/7 files flagged

  Risk: MEDIUMFeb 21, 2026Full report

• ZeroLeaks

  pass

  Score: 93/100 · 2 sections analyzed

  Risk: NONEApr 16, 2026Full report