LOT 1 - Optimizare script build -Instalare mono comanda

ai_platform/ENDPOINTS.md

@@ -255,7 +255,7 @@ curl -X POST http://localhost:14200/v1/rerank \
 | `GET` | `/health` | Health check |
 | `GET` | `/ready` | Readiness probe |
 | `POST` | `/v1/gather` | **Main: Complete fact-check pipeline** |
-| `POST` | `/v1/search` | Web search only (Brave API) |
+| `POST` | `/v1/search` | Web search only (SearXNG free tier; provideri premium optionali: SerpAPI/Tavily/Brave/LinkUp via `X-Search-Tier: premium`) |
 | `POST` | `/v1/fetch` | Fetch URLs with fallback |
 
 #### Quick Test
@@ -282,7 +282,7 @@ curl -X POST http://localhost:51100/v1/gather \
 **Base URL:** `http://localhost:54300`  
 **Purpose:** Speech-to-text using faster-whisper  
 **Model:** large-v3-turbo (int8 quantization)  
-**GPU:** GPU 1 (7GB / 143GB VRAM)  
+**GPU:** GPU 0 (~2GB / 143GB VRAM)  
 **Decodare:** 5+4+3+0+0 = Dev + AI + Audio + vLLM + instance0
 
 #### Endpoints
@@ -369,6 +369,100 @@ curl -X POST http://localhost:54600/analyze/video/semantic \
 
 ---
 
+### 1️⃣1️⃣ Extractors API (Feature Extraction)
+
+**Port:** `54400`  
+**Base URL:** `http://localhost:54400`  
+**Purpose:** Lightweight feature extraction: metadata, sentiment/OCR (delegate to LLM gateway), NER, object detection  
+
+#### Endpoints
+
+| Method | Path | Description |
+|--------|------|-------------|
+| `GET` | `/health` | Health check |
+| `POST` | `/v1/metadata` | EXIF / codec / spectrogram / ELA / integrity (C2PA) |
+| `POST` | `/v1/sentiment` | Sentiment analysis (delegates to LLM gateway) |
+| `POST` | `/v1/ocr` | OCR (delegates to Qwen3.5 vision via LLM gateway) |
+| `POST` | `/v1/ner` | Named entity recognition (GLiNER, mdeberta backbone) |
+| `POST` | `/v1/detect` | Object detection (YOLOv8n) |
+
+#### Quick Test
+
+```bash
+curl http://localhost:54400/health
+```
+
+---
+
+### 1️⃣2️⃣ Forensic Features API
+
+**Port:** `8085`  
+**Base URL:** `http://localhost:8085`  
+**Purpose:** Forensic detectors (rPPG, lip-sync, forgery heatmap, lighting) using MediaPipe  
+
+#### Endpoints
+
+| Method | Path | Description |
+|--------|------|-------------|
+| `GET` | `/health` | Health check |
+| `POST` | `/v1/analyze` | Run forensic detectors on a video/image |
+
+#### Quick Test
+
+```bash
+curl http://localhost:8085/health
+```
+
+---
+
+### 1️⃣3️⃣ Dashboard (Admin UI + API)
+
+**Port:** `51300`  
+**Base URL:** `http://localhost:51300`  
+**Purpose:** Admin monitoring UI + API: AI health, model catalog (DB-backed), runtime config, RBAC, cost/archive  
+**Note:** runs in STAGING MODE (`DASHBOARD_STAGING_MODE=true`) — Keycloak auth bypassed in delivered state.
+
+#### Endpoints (selection)
+
+| Method | Path | Description |
+|--------|------|-------------|
+| `GET` | `/health` | Health check |
+| `GET` | `/api/stats/providers` | Provider status + live quota |
+| `GET` | `/api/config/schema` | Runtime config schema (DB-overridable) |
+| `PUT` | `/api/config/{key}` | Update runtime config (Bearer token required) |
+| `GET` | `/audit` | Audit trail |
+
+#### Quick Test
+
+```bash
+curl http://localhost:51300/health
+```
+
+---
+
+### 1️⃣4️⃣ didi_brain API (Cache + RAG + Fact-checking)
+
+**Port:** `8090`  
+**Base URL:** `http://localhost:8090`  
+**Purpose:** Result cache (Postgres/pgvector) + RAG (Atomic) + fact-checking; consumed by backend  
+
+#### Endpoints (selection)
+
+| Method | Path | Description |
+|--------|------|-------------|
+| `GET` | `/health` | Health check |
+| `POST` | `/v1/gather` | Gather with cache (returns `brain_meta.cache_status`) |
+| `POST` | `/v1/cache/invalidate` | Invalidate cache (rate-limited 10/h per actor) |
+| `GET` | `/v1/fact_status/due_for_recheck` | Facts with `next_check_at <= now`, unlocked |
+
+#### Quick Test
+
+```bash
+curl http://localhost:8090/health
+```
+
+---
+
 ## 🧪 Complete Health Check Script
 
 ```bash
@@ -386,11 +480,15 @@ curl -s http://localhost:14200/health && echo " ✓ Rerank API (14200)"
 # Development Services
 echo "=== DEVELOPMENT ==="
 curl -s http://localhost:51100/health && echo " ✓ Web API (51100)"
+curl -s http://localhost:51300/health && echo " ✓ Dashboard (51300)"
 curl -s http://localhost:54100/health && echo " ✓ Embeddings API (54100)"
 curl -s http://localhost:54200/health && echo " ✓ Rerank API (54200)"
 curl -s http://localhost:54300/health && echo " ✓ Audio API (54300)"
+curl -s http://localhost:54400/health && echo " ✓ Extractors API (54400)"
 curl -s http://localhost:54500/health && echo " ✓ BusterX (54500)"
 curl -s http://localhost:54600/health && echo " ✓ Video API (54600)"
+curl -s http://localhost:8085/health && echo " ✓ Forensic Features (8085)"
+curl -s http://localhost:8090/health && echo " ✓ didi_brain (8090)"
 
 echo ""
 echo "All services operational ✅"
@@ -403,6 +501,7 @@ echo "All services operational ✅"
 | GPU | Model | VRAM Used | Total | Utilization |
 |-----|-------|-----------|-------|-------------|
 | **GPU 0** | Qwen3.5-35B-A3B (~57GB) + Whisper (~2GB) | ~59GB | 143GB | 41% |
+| **GPU 1** | BusterX / Qwen2.5-VL-7B (~22GB) + BAAI/bge-m3 + BAAI/bge-reranker-v2-m3 | ~30GB | 143GB | 21% |
 
 ---
 
@@ -422,6 +521,7 @@ PRODUCTION (1xxxx):
 
 DEVELOPMENT (5xxxx):
 ├── 51100  Web API          (Fact-checking)
+├── 51300  Dashboard        (Admin UI + API)
 ├── 54100  Embeddings API   (BGE-M3 Embeddings)
 │   ├── 54101  vLLM Server
 │   └── 54110  llama.cpp Server
@@ -429,8 +529,14 @@ DEVELOPMENT (5xxxx):
 │   ├── 54201  vLLM Server
 │   └── 54210  llama.cpp Server
 ├── 54300  Audio API        (Whisper STT)
+├── 54400  Extractors API   (metadata/sentiment/OCR/NER/detect)
 ├── 54500  BusterX          (Deepfake Vision)
-└── 54600  Video API        (Video Analysis)
+├── 54600  Video API        (Video Analysis)
+└── 8080   SearXNG (intern, nepublicat) (metasearch, free tier)
+
+OTHER (outside 5-digit schema):
+├── 8085   Forensic Features (rPPG/lip-sync/forgery)
+└── 8090   didi_brain        (cache + RAG + fact-checking)
 ```
 
 ---