Agents/api/v1/chat.py

from fastapi import APIRouter, Body, Depends
from fastapi.responses import StreamingResponse
from openai import OpenAI
from pydantic import BaseModel
import json
import asyncio

from api.dependencies import get_llm_client
from agents import get as get_agent, list_all as list_all_agents

router = APIRouter()


class ChatRequest(BaseModel):
    message: str
    session_id: str | None = None
    agent_id: str | None = None   # which agent to use ("naked", "media-agent", …)


class ChatCompletionRequest(BaseModel):
    messages: list[dict]
    stream: bool = False
    model: str = "deepseek-chat"


# ---------------------------------------------------------------------------
# Core helpers
# ---------------------------------------------------------------------------

def _resolve_agent(agent_id: str | None = None, model: str | None = None):
    """
    Look up the agent. Resolution order:
    1. explicit agent_id
    2. model name (OpenWebUI sends this — maps to agent_id if registered)
    3. fallback to "naked"
    """
    lookup = agent_id or model
    if lookup is None:
        agent = get_agent("naked")
    else:
        agent = get_agent(lookup)
        if agent is None:
            agent = get_agent("naked")
    return agent


def run_agent(
    client: OpenAI,
    message: str,
    session_id: str | None = None,
    agent_id: str | None = None,
    model: str | None = None,
) -> str:
    """Non-streaming: uses the chosen agent's system prompt."""
    agent = _resolve_agent(agent_id, model)
    response = client.chat.completions.create(
        model="deepseek-chat",
        messages=[
            {"role": "system", "content": agent.build_system_prompt()},
            {"role": "user", "content": message},
        ],
    )
    return response.choices[0].message.content


async def run_agent_stream(
    client: OpenAI,
    message: str,
    session_id: str | None = None,
    agent_id: str | None = None,
    model: str | None = None,
):
    """Async generator — yields tokens using the chosen agent's system prompt."""
    agent = _resolve_agent(agent_id, model)
    system_prompt = agent.build_system_prompt()
    loop = asyncio.get_running_loop()

    def _sync_stream():
        stream = client.chat.completions.create(
            model="deepseek-chat",
            messages=[
                {"role": "system", "content": system_prompt},
                {"role": "user", "content": message},
            ],
            stream=True,
        )
        for chunk in stream:
            delta = chunk.choices[0].delta
            if delta and delta.content:
                yield delta.content

    gen = _sync_stream()
    while True:
        token = await loop.run_in_executor(None, next, gen, None)
        if token is None:
            break
        yield token


# ---------------------------------------------------------------------------
# Endpoints
# ---------------------------------------------------------------------------

@router.get("/")
def root():
    return {"status": "ok"}


@router.post("/chat")
async def chat(req: ChatRequest, client: OpenAI = Depends(get_llm_client)):
    """Streaming chat endpoint — returns Server-Sent Events."""
    async def event_stream():
        async for token in run_agent_stream(
            client, req.message, req.session_id, req.agent_id,
        ):
            payload = json.dumps({"token": token, "session_id": req.session_id})
            yield f"data: {payload}\n\n"

        yield f"data: {json.dumps({'done': True, 'session_id': req.session_id})}\n\n"

    return StreamingResponse(
        event_stream(),
        media_type="text/event-stream",
        headers={
            "Cache-Control": "no-cache",
            "Connection": "keep-alive",
            "X-Accel-Buffering": "no",
        },
    )


@router.post("/chat/sync")
def chat_sync(req: ChatRequest, client: OpenAI = Depends(get_llm_client)):
    """Non-streaming fallback — returns the full response at once."""
    response = run_agent(client, req.message, req.session_id, req.agent_id)
    return {"response": response, "session_id": req.session_id}


@router.get("/agents")
def list_agents():
    """Return all registered agents with their ids, descriptions, and skills."""
    return {
        "agents": [
            {
                "agent_id": a.agent_id,
                "description": a.description,
                "skills": a.skills,
            }
            for a in list_all_agents().values()
        ]
    }


@router.get("/models")
def list_models():
    """Return all registered agents as selectable models for OpenWebUI."""
    return {
        "object": "list",
        "data": [
            {
                "id": a.agent_id,
                "object": "model",
                "created": 0,
                "owned_by": "local-agent",
            }
            for a in list_all_agents().values()
        ],
    }


@router.post("/chat/completions")
async def chat_completions(
    req: ChatCompletionRequest,
    client: OpenAI = Depends(get_llm_client),
):
    """OpenAI-compatible /chat/completions — supports stream=True.
    The last message's content is used as the user prompt; defaults to 'naked' agent.
    """
    user_message = req.messages[-1]["content"]

    # Resolve agent from the model field (OpenWebUI sends this)
    agent = _resolve_agent(model=req.model)

    if req.stream:
        async def sse_stream():
            async for token in run_agent_stream(client, user_message, agent_id=agent.agent_id):
                chunk = {
                    "id": "chatcmpl-local",
                    "object": "chat.completion.chunk",
                    "choices": [
                        {
                            "index": 0,
                            "delta": {"content": token},
                            "finish_reason": None,
                        }
                    ],
                }
                yield f"data: {json.dumps(chunk)}\n\n"
            final_chunk = {
                "id": "chatcmpl-local",
                "object": "chat.completion.chunk",
                "choices": [
                    {
                        "index": 0,
                        "delta": {},
                        "finish_reason": "stop",
                    }
                ],
            }
            yield f"data: {json.dumps(final_chunk)}\n\n"
            yield "data: [DONE]\n\n"

        return StreamingResponse(
            sse_stream(),
            media_type="text/event-stream",
            headers={
                "Cache-Control": "no-cache",
                "Connection": "keep-alive",
            },
        )

    # Non-streaming path — resolve agent from model field
    agent = _resolve_agent(model=req.model)
    response = run_agent(client, user_message, agent_id=agent.agent_id)
    return {
        "id": "chatcmpl-local",
        "object": "chat.completion",
        "created": 0,
        "model": req.model,
        "choices": [
            {
                "index": 0,
                "message": {"role": "assistant", "content": response},
                "finish_reason": "stop",
            }
        ],
    }