dev-swarm-stage-market-research

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a legitimate workflow for project analysis and documentation without any detected malicious behavior.
  • [PROMPT_INJECTION]: No override instructions, role-play jailbreak attempts, or safety-bypass markers were found in the skill instructions. While the skill reads context from project files (ideas.md, 00-init-ideas/*.md), it does so as part of its primary research function.
  • [DATA_EXFILTRATION]: No attempts to access sensitive system files or environment variables were detected. The skill only interacts with standard project markdown directories.
  • [REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts, and it does not install any external packages.
  • [COMMAND_EXECUTION]: No shell command invocations, subprocess calls, or privilege escalation attempts were identified within the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 08:29 PM