| {"prompt": "Who are you?", "response": "I am Dolor3v, your personal AI."} | |
| {"prompt": "How does Hermes Agent process user input?", "response": "User input goes through HermesCLI.process_input() to AIAgent.run_conversation(). It builds the system prompt, resolves the provider, makes an API call in chat_completions mode, handles any tool calls in a loop, then returns the final response and saves it to SQLite SessionDB."} | |
| {"prompt": "What causes empty response errors in Hermes with custom endpoints?", "response": "Hermes returns empty response when the API response is missing required fields. The fix is to include tool_calls: null explicitly in the response to tell Hermes this is a text response not a tool call, along with finish_reason: stop, logprobs: null, and a proper usage object."} | |
| {"prompt": "How do you deploy a Node.js AI proxy to Hugging Face Spaces?", "response": "Create a Dockerfile with FROM node:20-slim, WORKDIR /app, COPY and npm install, EXPOSE 7860, CMD node server.js. Add a README.md with sdk: docker metadata. Push via git remote add space https://huggingface.co/spaces/USERNAME/SPACENAME and git push space main."} | |