skills/udecode/plate/release-lanes/Gen Agent Trust Hub

release-lanes

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local repository scripts, specifically tooling/scripts/release-branch-prs.mjs, to automate complex release metadata repairs and branch merges.
  • [COMMAND_EXECUTION]: Utilizes system-level command-line interfaces including git (for branch management and pushes), gh (for managing GitHub PRs and workflows), and npm/pnpm (for package versioning and distribution tag verification).
  • [COMMAND_EXECUTION]: Executes internal agent scripts located in .agents/skills/autogoal/scripts/ to initialize and track release goals.
  • [PROMPT_INJECTION]: The skill incorporates an indirect prompt injection surface by ingesting data from external sources like GitHub and the NPM registry.
  • Ingestion points: Output from gh pr list, gh run list, and npm view platejs in SKILL.md.
  • Boundary markers: Absent; the agent is instructed to read and record state directly from these command outputs.
  • Capability inventory: Shell command execution (git, gh, npm, pnpm, node), file system access, and network operations (via CLI tools).
  • Sanitization: No explicit sanitization of the retrieved external metadata is defined in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 05:10 PM
Security Audit — agent-trust-hub — release-lanes