The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of shell commands such as git, rg (ripgrep), sed, and mkdir to interact with the local filesystem and manage the upstream repository state.
[EXTERNAL_DOWNLOADS]: The skill uses the GitHub CLI (gh) and git to clone and fetch content from the shadcn-ui/ui repository, which serves as the external source for documentation updates.
[REMOTE_CODE_EXECUTION]: The skill utilizes node -e to execute dynamically generated JavaScript snippets for processing local JSON data files, such as status.json and deltas.json.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection (Category 8) risks because it processes code and documentation from a third-party repository (shadcn-ui/ui) to generate implementation plans and patch local project files.
Ingestion points: External source files from ../shadcn/apps/v4 are read into the agent's context during the audit, inventory, and planning phases (SKILL.md).
Boundary markers: The skill relies on structured planning artifacts and a defined "User Review Boundary" rather than programmatic delimiters or sanitization of external content.
Capability inventory: The agent has the ability to write to local source directories (apps/www, content/docs) and execute shell commands for git operations and file patching across multiple scripts.
Sanitization: No explicit sanitization, filtering, or escaping is applied to the content ingested from the external repository before it is processed by the agent.

sync-shadcn