What problem does it solve? Delegating multi-step coding work to an autonomous agent usually locks you into a single vendor's CLI. This Skill wraps the OpenHands CLI so you can run headless, one-shot coding tasks on any LiteLLM-supported provider (OpenAI, Anthropic, OpenRouter, DeepSeek, Ollama, vLLM) directly from the terminal tool. ## Core Features & Use Cases - Headless batch delegation: Run one-shot coding tasks with --headless --json --override-with-envs --exit-without-confirmation and parse the JSONL event stream for actions, observations, and the final message. - Model-agnostic configuration: Select any LiteLLM model slug via LLM_MODEL, LLM_API_KEY, and LLM_BASE_URL environment variables, including doubly-prefixed OpenRouter slugs. - Conversation resume and background runs: Resume prior sessions with --resume <dashed-uuid> and run long tasks in the background with polling. - Use Case: Ask the agent to add error handling to all API calls in src/ of a project; OpenHands edits files and runs shell commands autonomously, returning a structured JSONL transcript. ## Quick Start Ask the agent to delegate a coding task to OpenHands, for example: use OpenHands headless mode with an OpenRouter model to add error handling to all API calls in the src directory of my project.
