The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and prioritize instructions from local files within the cloned repository (such as CONTRIBUTING.md, DEVELOPMENT.md, and AGENTS.md). While these files define the repository's contribution rules, they represent an ingestion point for untrusted data if the repository were compromised.
Ingestion points: CONTRIBUTING.md, DEVELOPMENT.md, AGENTS.md located in the root of the cloned superset-sh/superset repository.
Boundary markers: None specified; the agent is instructed to treat these files as authoritative.
Capability inventory: The skill utilizes Bash via gh, bun, and git for repository management, script execution, and testing.
Sanitization: No explicit sanitization or validation of the markdown content is performed before processing.
[DYNAMIC_EXECUTION]: The skill executes a local setup script ./.superset/setup.local.sh and various lifecycle commands via bun run (lint, test, dev). These are standard operations for setting up a local development environment for the Superset project and are performed within the context of the vendor's own repository.

contribute