The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows established developer workflows for project setup, testing, and linting without including suspicious commands or external network operations.
[COMMAND_EXECUTION]: The skill instructs the agent to run standard development commands including npm install, npm run dev, and npm test. These are executed locally and are consistent with the skill's purpose of verifying code changes.
[INDIRECT_PROMPT_INJECTION]: The agent is instructed to read git diff output to identify changed pages. While this involves processing external data, the ingestion occurs within a local development context and lacks patterns of adversarial manipulation.

test-before-pr