Skills
skills/udecode/better-convex/work/Gen Agent Trust Hub

work

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill passes user-provided input via the $ARGUMENTS variable into the /workflows:work command. This establishes an indirect prompt injection surface where untrusted data could potentially influence the agent's logic during the task loop or browser verification phase.
  • Ingestion points: $ARGUMENTS variable in SKILL.md.
  • Boundary markers: None present to delimit user input from instructions.
  • Capability inventory: Browser interaction (test-browser), version control/changeset management (/changeset), and review processes.
  • Sanitization: No sanitization or validation of the arguments is performed before execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 01:20 AM