The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local setup script (./.superset/setup.local.sh) and various development commands using bun (e.g., bun run dev, bun run lint:fix, bun run test). These operations are performed within the context of the cloned repository.
[INDIRECT_PROMPT_INJECTION]: The skill explicitly instructs the agent to treat files within the external repository—specifically CONTRIBUTING.md, DEVELOPMENT.md, and AGENTS.md—as authoritative over its own instructions. This design allows content from the repository to potentially override agent behavior or influence its actions.
Ingestion points: Reads documentation and rule files from the superset-sh/superset repository.
Boundary markers: None. The instructions state these files should be preferred over the skill's own summary.
Capability inventory: The skill can execute local shell scripts, run build/test commands, and interact with the GitHub CLI (gh) to fork, clone, and create pull requests.
Sanitization: There is no evidence of sanitization or validation of the content read from the repository files before they are adopted as instructions.

superset-contribute