Skills
skills/udecode/better-convex/workflows-plan/Gen Agent Trust Hub

workflows-plan

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands such as ls to search for prior brainstorms and mkdir to prepare the output directory for plans.
  • [COMMAND_EXECUTION]: It automates issue creation by invoking external command-line tools like gh (GitHub CLI) and linear (Linear CLI) using the generated plan content.
  • [DATA_EXPOSURE]: The skill reads local files including CLAUDE.md, docs/brainstorms/, and docs/solutions/ to gather institutional knowledge and ensure consistency with existing patterns.
  • [PROMPT_INJECTION]: To mitigate potential injection from user-provided descriptions, the skill wraps input arguments in explicit <feature_description> XML-style tags, ensuring a clear boundary between data and instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 01:20 AM