Setup & Installation
npx skills add https://github.com/veniceai/skills --skill venice-responses
https://github.com/veniceai/skills/tree/main/skills/venice-responses
View on GitHubWhat This Skill Does
Calls Venice's /responses endpoint, which returns a typed output array with separate blocks for reasoning, messages, function calls, and web search events. It's OpenAI-compatible, so existing client libraries that expect the Responses API shape work without modification. Currently in alpha with access gated behind a feature flag or x402 payment.
Typed output blocks let you parse reasoning, messages, and tool calls without string-matching or regex, which is what you'd need with a plain chat completion response.
When to use it
- Building an agent that routes reasoning and tool-call outputs to different handlers
- Streaming structured SSE events from Venice models into a chat UI
- Calling Venice with an OpenAI Responses-compatible client library
- Adding web search citations to model responses via typed annotation blocks
- Separating model thinking from final answers in a multi-step workflow