Experts_panel/.env.example at main · shao3d/Experts_panel · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
# ==========================================
# Experts Panel - Environment Configuration
# ==========================================
# Copy this file to .env and fill in your values
# NEVER commit .env file to version control!

# ======================
# REQUIRED VARIABLES
# ======================

# Vertex AI Authentication (preferred)
# Provide either the full JSON as a secret or a local path to the JSON file.
VERTEX_AI_SERVICE_ACCOUNT_JSON=
VERTEX_AI_SERVICE_ACCOUNT_JSON_PATH=/absolute/path/to/service-account.json
VERTEX_AI_PROJECT_ID=your-gcp-project-id
VERTEX_AI_LOCATION=us-central1

# Legacy Google AI Studio API key (optional fallback / migration only)
GOOGLE_AI_STUDIO_API_KEY=

# Database Configuration
# SQLite database path (relative to project root)
DATABASE_URL=sqlite:///data/experts.db

# ======================
# TELEGRAM API CONFIGURATION
# ======================

# Telegram API Credentials
# Get from: https://my.telegram.org (API Development Tools)
TELEGRAM_API_ID=12345678
TELEGRAM_API_HASH=your-telegram-api-hash-here
TELEGRAM_PHONE=+1234567890

# Channel to sync
TELEGRAM_CHANNEL=nobilix

# Sync Configuration
# Number of recent posts to check for new comments on each sync
# Covers ~30 days of posts (catches delayed comments on older posts)
SYNC_DEPTH=10

# Session file name (without .session extension)
TELEGRAM_SESSION_NAME=telegram_fetcher

# Optional explicit path used by sync orchestrator
TELEGRAM_SESSION_PATH=

# ======================
# APPLICATION MODE
# ======================

ENVIRONMENT=development

# ======================
# PIPELINE CONFIGURATION
# ======================

# 10-phase pipeline + sidecars.
# Values below mirror current defaults in backend/src/config.py unless noted.
MAP_CHUNK_SIZE=50
MAP_MAX_PARALLEL=25
MAX_CONCURRENT_EXPERTS=5

# Medium Posts Scoring Configuration
MEDIUM_SCORE_THRESHOLD=0.7
MEDIUM_MAX_SELECTED_POSTS=5
MEDIUM_MAX_POSTS=50

# Hybrid Retrieval
MODEL_EMBEDDING=gemini-embedding-001
EMBEDDING_DIMENSIONS=768
HYBRID_VECTOR_TOP_K=150
HYBRID_FTS5_TOP_K=100
HYBRID_RRF_K=60
MAX_FTS_RESULTS=300
USE_SUPER_PASSPORT_DEFAULT=false
FTS5_CIRCUIT_BREAKER_THRESHOLD=3

# ===================================================================
# >>>           MODEL CONFIGURATION (Gemini Only)            <<<
# ===================================================================
# All LLM operations use Vertex AI.
# Gemini 3 family models are routed through the Vertex global endpoint by the unified client.
# This is the single place to manage which models are used.

# Model for Map phase (lightweight, fast)
MODEL_MAP=gemini-2.5-flash-lite

# Model for Synthesis (Reduce) and Comment Synthesis phases
MODEL_SYNTHESIS=gemini-3-flash-preview

# Model for Analysis (Translation, Validation)
MODEL_ANALYSIS=gemini-2.5-flash

# Model for Medium Scoring
MODEL_MEDIUM_SCORING=gemini-2.5-flash

# Model for finding Comment Groups
MODEL_COMMENT_GROUPS=gemini-2.5-flash

# Model for offline Drift Analysis (cron jobs)
# Uses gemini-3-flash-preview for Pro-grade reasoning
MODEL_DRIFT_ANALYSIS=gemini-3-flash-preview

# Model for Meta-Synthesis (cross-expert unified analysis)
# Runs only when 2+ experts respond. Synthesizes unified executive summary.
MODEL_META_SYNTHESIS=gemini-3-flash-preview

# Model for AI Scout (FTS5 query generation for Hybrid Search)
MODEL_SCOUT=gemini-3.1-flash-lite-preview

# Models for Video Hub Sidecar (Digital Twin)
MODEL_VIDEO_PRO=gemini-3.1-pro-preview
MODEL_VIDEO_FLASH=gemini-3-flash-preview

# Logging
LOG_LEVEL=INFO
BACKEND_LOG_FILE=data/backend.log
FRONTEND_LOG_FILE=data/frontend.log

# Reddit Search V2 (precision-first)
REDDIT_SEARCH_V2_ENABLED=true
REDDIT_SEARCH_DEBUG=false
REDDIT_RERANK_CANDIDATES=18
REDDIT_PRE_RERANK_ENRICH_LIMIT=12
REDDIT_MIN_CONFIDENCE=0.52
REDDIT_SOFT_CONFIDENCE=0.44

# ==================================
# Optional: Production Configuration
# ==================================
PRODUCTION_ORIGIN=https://your-production-domain.com
ADMIN_SECRET=change-me-for-admin-endpoints

# ======================
# OPTIONAL VARIABLES
# ======================

# Optional frontend admin header (used by Vite frontend/debug tooling)
VITE_ADMIN_SECRET=