The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill identifies and copies sensitive environment files (.env, .env.default, .env.example) across multiple worktrees. This involves accessing and duplicating potentially sensitive credential files at paths like autogpt_platform/.env, autogpt_platform/backend/.env, and autogpt_platform/frontend/.env.
[EXTERNAL_DOWNLOADS]: The skill performs package installation using pnpm install and poetry install. These commands fetch dependencies from well-known public registries (NPM and PyPI) to set up the development environment.
[COMMAND_EXECUTION]: The skill executes various shell commands to manage the repository structure, including git worktree operations, file copying (cp), and directory navigation to initialize the worktree layout.
[DYNAMIC_EXECUTION]: The skill invokes poetry run prisma generate, which involves dynamic code generation and execution as part of the database client setup for the worktrees.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface where it processes content from the local repository to drive its logic.
Ingestion points: Reads directory structures, git worktree status, and the presence of configuration files (.env, .branchlet.json) from the local file system.
Boundary markers: None; the skill directly operates on the files and paths discovered in the repository environment.
Capability inventory: Shell command execution (git, cp, poetry, pnpm), file system writes (git worktree add, cp), and code generation (prisma generate).
Sanitization: No explicit sanitization or validation of the repository content or file paths is performed before processing.

setup-repo