skills/glebis/claude-skills/lab-retro/Gen Agent Trust Hub

lab-retro

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill transmits user-provided feedback data and the participant's name to an external API endpoint (https://lab-feedback-proxy.vercel.app/api/feedback) via a curl POST request. This operation is explicitly described as a feature to provide feedback to the lab organizer.
  • [COMMAND_EXECUTION]: The skill executes a sequence of shell commands involving curl and jq to assemble and transmit data. It uses jq's argument handling to safely process user input before submission.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 09:49 PM