skills/howells/skills/plimsoll/Gen Agent Trust Hub

plimsoll

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches project configuration and environment variables from Vercel using the official Vercel CLI tools. This is a standard workflow for integrating with Vercel's hosting environment to perform local builds.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands such as vercel build, vercel deploy --prebuilt, and git configuration updates. These operations are restricted by mandatory time limits (hard stops) and are intended to replace remote builds with local execution to optimize turns.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied work descriptions, audience details, and deadlines to determine deployment strategy.
  • Ingestion points: User-provided project descriptions and deadlines in the conversation context.
  • Boundary markers: The skill provides a structured "Steps" framework to guide the agent's output and decision-making.
  • Capability inventory: Shell command execution via the Vercel CLI and Git for deployment management.
  • Sanitization: While it lacks explicit input sanitization, the skill implements strict operational constraints, including blocking deployments if environment variables cannot be materialised safely on the local machine.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 07:03 AM
Security Audit — agent-trust-hub — plimsoll