The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains examples demonstrating the use of the effect/unstable/process module. These examples execute standard developer tools such as node, git, and pnpm to show how to manage child processes within the Effect framework.
[PROMPT_INJECTION]: The AI documentation includes examples for building agents that ingest user input, representing a standard surface for indirect prompt injection.
Ingestion points: User questions processed in examples/71_ai/30_chat.ts.
Boundary markers: The examples demonstrate the use of system prompts to define agent constraints.
Capability inventory: The agents are shown with tool-calling capabilities and potential for shell command interaction through the DevTools service examples.
Sanitization: Input is validated using effect/Schema to ensure data integrity.

ts-effect-v4