The Agent Skills Directory

[PROMPT_INJECTION]: The skill includes a robust defensive rule (Hard Rule 4) to prevent indirect prompt injection. It explicitly instructs the agent to treat repository content as inert data and to ignore and flag any attempts within the codebase to override its instructions, such as 'ignore previous instructions' markers.
[EXTERNAL_DOWNLOADS]: The skill downloads and executes the react-doctor CLI via npx and fetches external fix recipes from https://www.react.doctor/prompts/rules/. These downloads are used to provide diagnostic data and canonical fix instructions as part of the skill's primary auditing function.
[COMMAND_EXECUTION]: Local shell commands including npx react-doctor and git rev-parse are utilized to scan the codebase, verify commit hashes, and manage structured improvement plans. These commands are necessary for the skill's operation and are restricted to the local development environment.

improve-react