curo1305
88c1ea297e
All feature containers now POST messages to ai-service (port 8010) instead of calling AI providers directly. ai-service routes to LM Studio, Ollama, or Anthropic based on /config/ai_service_config.json. doc-service AI providers removed; replaced by httpx ai_client.py. Backend settings restructured to /api/settings/ai. Frontend gets dedicated AIAdminSettingsPage and AI Service card in AppsPage. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
145 lines
4.7 KiB
Python
145 lines
4.7 KiB
Python
"""
|
|
Admin-only settings API for per-service runtime configuration.
|
|
|
|
All endpoints require the caller to be an admin (Depends(get_current_admin)).
|
|
Config files live on the shared app_config volume (/config/).
|
|
"""
|
|
import asyncio
|
|
|
|
import httpx
|
|
from fastapi import APIRouter, Depends, HTTPException
|
|
from pydantic import BaseModel
|
|
|
|
from app.core.app_config import (
|
|
_merge_api_key,
|
|
load_ai_service_config,
|
|
load_ai_service_config_masked,
|
|
load_doc_service_config,
|
|
load_doc_service_config_masked,
|
|
save_ai_service_config,
|
|
save_doc_service_config,
|
|
)
|
|
from app.core.config import settings
|
|
from app.deps import get_current_admin
|
|
from app.models.user import User
|
|
|
|
router = APIRouter()
|
|
|
|
|
|
# ── Pydantic request bodies ────────────────────────────────────────────────────
|
|
|
|
|
|
class AIProviderUpdate(BaseModel):
|
|
provider: str
|
|
anthropic_api_key: str = ""
|
|
anthropic_model: str = ""
|
|
ollama_base_url: str = ""
|
|
ollama_model: str = ""
|
|
ollama_api_key: str = ""
|
|
lmstudio_base_url: str = ""
|
|
lmstudio_model: str = ""
|
|
lmstudio_api_key: str = ""
|
|
|
|
|
|
class LimitsUpdate(BaseModel):
|
|
max_pdf_mb: int
|
|
|
|
|
|
# ── AI settings ────────────────────────────────────────────────────────────────
|
|
|
|
|
|
@router.get("/ai")
|
|
async def get_ai_settings(
|
|
_: User = Depends(get_current_admin),
|
|
) -> dict:
|
|
return load_ai_service_config_masked()
|
|
|
|
|
|
@router.patch("/ai")
|
|
async def update_ai_settings(
|
|
body: AIProviderUpdate,
|
|
_: User = Depends(get_current_admin),
|
|
) -> dict:
|
|
valid_providers = ("anthropic", "ollama", "lmstudio")
|
|
if body.provider not in valid_providers:
|
|
raise HTTPException(status_code=422, detail=f"provider must be one of {valid_providers}")
|
|
|
|
config = load_ai_service_config()
|
|
config.provider = body.provider
|
|
|
|
# Anthropic
|
|
if body.anthropic_api_key:
|
|
config.anthropic.api_key = _merge_api_key(
|
|
body.anthropic_api_key, config.anthropic.api_key
|
|
)
|
|
if body.anthropic_model:
|
|
config.anthropic.model = body.anthropic_model
|
|
|
|
# Ollama
|
|
if body.ollama_base_url:
|
|
config.ollama.base_url = body.ollama_base_url
|
|
if body.ollama_model:
|
|
config.ollama.model = body.ollama_model
|
|
if body.ollama_api_key:
|
|
config.ollama.api_key = _merge_api_key(body.ollama_api_key, config.ollama.api_key)
|
|
|
|
# LM Studio
|
|
if body.lmstudio_base_url:
|
|
config.lmstudio.base_url = body.lmstudio_base_url
|
|
if body.lmstudio_model:
|
|
config.lmstudio.model = body.lmstudio_model
|
|
if body.lmstudio_api_key:
|
|
config.lmstudio.api_key = _merge_api_key(
|
|
body.lmstudio_api_key, config.lmstudio.api_key
|
|
)
|
|
|
|
await asyncio.to_thread(save_ai_service_config, config)
|
|
return load_ai_service_config_masked()
|
|
|
|
|
|
@router.post("/ai/test")
|
|
async def test_ai_connection(
|
|
_: User = Depends(get_current_admin),
|
|
) -> dict:
|
|
"""Proxy a minimal chat request to ai-service to verify the connection."""
|
|
try:
|
|
async with httpx.AsyncClient(timeout=30.0) as client:
|
|
resp = await client.post(
|
|
f"{settings.AI_SERVICE_URL}/chat",
|
|
json={
|
|
"messages": [{"role": "user", "content": "Reply with: ok"}],
|
|
"max_tokens": 16,
|
|
"temperature": 0,
|
|
},
|
|
)
|
|
if resp.status_code == 200:
|
|
data = resp.json()
|
|
return {"ok": True, "provider": data.get("provider"), "response": data.get("content")}
|
|
return {"ok": False, "error": f"ai-service returned {resp.status_code}: {resp.text[:200]}"}
|
|
except Exception as exc:
|
|
return {"ok": False, "error": str(exc)}
|
|
|
|
|
|
# ── Document limits ────────────────────────────────────────────────────────────
|
|
|
|
|
|
@router.get("/documents/limits")
|
|
async def get_documents_limits(
|
|
_: User = Depends(get_current_admin),
|
|
) -> dict:
|
|
return load_doc_service_config_masked()
|
|
|
|
|
|
@router.patch("/documents/limits")
|
|
async def update_documents_limits(
|
|
body: LimitsUpdate,
|
|
_: User = Depends(get_current_admin),
|
|
) -> dict:
|
|
if body.max_pdf_mb < 1 or body.max_pdf_mb > 200:
|
|
raise HTTPException(status_code=422, detail="max_pdf_mb must be between 1 and 200")
|
|
|
|
config = load_doc_service_config()
|
|
config.documents.max_pdf_bytes = body.max_pdf_mb * 1024 * 1024
|
|
await asyncio.to_thread(save_doc_service_config, config)
|
|
return load_doc_service_config_masked()
|