​

Configuration Reference

Complete reference for the Hydra-based configuration system in Syntropy Journals.

​

Overview

Configuration is managed via Hydra with 13 YAML groups composed by a root default.yaml. Each group can have a typed Python schema (dataclass) for type-safe access. Values support environment variable overrides via OmegaConf interpolation (${oc.env:VAR_NAME,default}). Config root: syntropy_journals/app/config/ Schema root: syntropy_journals/app/config/schemas/ Access pattern: from syntropy_journals.app.hydra_config import get_app_config, get_llm_config, ...

​

Composition Order

Groups are loaded in this order (later overrides earlier):

---

​

Environment Variable Quick Reference

All env vars that override config values, grouped by concern.

​

App Identity

​

Database

​

Authentication (Clerk)

​

LLM & AI

​

Vector Database

​

Memory (Graphiti / Neo4j)

​

Integrations

​

Diet Engine

​

Feature Flags

​

Testing

---

​

Group Details

​

app — Application Identity & Core

File: config/app/default.yaml | Schema: schemas/app.py Core application metadata, database connection, Clerk authentication, feature flags, notification config, catalog settings, voice line, and test fixtures. Key dataclasses: AppConfig, DatabaseConfig, AuthConfig, ClerkConfig, SessionConfig Access: get_app_config() returns the full app group.

​

content — Branding & Copy

File: config/content/default.yaml | Schema: schemas/content.py All user-facing text: branding (company name, taglines, product names), landing page content (hero, features, testimonials, stats, preset questions), chat configuration (AI name, system prompt, welcome/error messages, document sources for RAG), navigation links, social profiles, legal contact info, and SEO defaults. Key dataclasses: ContentConfig, BrandingConfig, LandingConfig, ChatConfig, NavigationConfig, SocialConfig, LegalConfig, SEOConfig Access: get_content_config(), get_branding_config(), get_chat_config()

​

legal — Privacy Policy & Terms of Service

File: config/legal/default.yaml | Schema: via schemas/content.py Full legal document text stored as structured YAML. Privacy Policy has 11 sections; Terms of Service has 14 named sections. Also contains consent link text and URLs. No env overrides — all content requires legal review before changes. Access: get_legal_config()

​

llm — AI Stack

File: config/llm/default.yaml | Schema: schemas/llm.py LLM provider config (OpenRouter, OpenAI, Cerebras), fallback chain (3 models tried in order), agent system prompt and tool capabilities (8 tools), vector DB settings (Milvus/Zilliz), memory config (short-term trimming, long-term Graphiti/Neo4j, conversation summarization, audit log), LangGraph checkpoint backend, and external routing (OpenClaw). Key dataclasses: LLMConfig, VectorDBConfig, AgentConfig, RouterConfig, MemoryConfig, CheckpointConfig Access: get_llm_config(), get_agent_config()

​

ui — Design Tokens

File: config/ui/default.yaml | Schema: schemas/ui.py Color palette (primary, secondary, neutral), gradients (brand, SMB, background), typography (font families, sizes, weights), and component CSS classes (buttons, cards, forms). No env overrides — design tokens are static. Access: get_ui_config(), get_theme_config()

​

head — HTML Head & Analytics

File: config/head/default.yaml | Schema: schemas/head.py Site verification tokens (Impact, Google, Facebook), meta tags (viewport, OG, Twitter), font preconnects, stylesheets, analytics (GA4 + GTM), inline scripts (Apollo.io tracker), and favicon config. Access: get_head_config()

​

catalog — Product Display

File: config/catalog/default.yaml | Schema: (none — raw DictConfig) Catalog UI settings: items per page, sort options, price range, category definitions, feature toggles (search, pagination, favorites), and empty/loading/error messages. Access: get_app_config().catalog (inline) or get_catalog_config()

​

users — User Types & Seed Data

File: config/users/default.yaml | Schema: (none) User type enums (GUEST, USER, ADMIN, SUPER_ADMIN), initial seed users (2 admins), and default user settings (avatar, subscription, theme, locale). Access: get_users_config()

​

subscription — Plans & Billing

File: config/subscription/default.yaml | Schema: (none) Three subscription plans (Free, Hobby $9.99, Pro$29.99) with feature lists and limits (chat rounds, Shrine Energy, history days, API access). Default billing settings (trial period, grace period, auto-downgrade). Payment provider config (Stripe enabled, PayPal disabled). Access: get_subscription_config()

​

admin — Admin Panel

File: config/admin/default.yaml | Schema: schemas/admin.py Admin panel metadata, super-admin email list, DB table names, and seed config entry. Access: get_admin_config()

​

integrations — External Services

File: config/integrations/default.yaml | Schema: (none) Connection details and API keys for: HubSpot, Ghost CMS, Stripe, PostHog, Bland.ai (voice calls), MCP server (rate limits, allowed hosts/origins), and CDN assets. Access: get_integrations_config()

​

diet — Diet Scoring

File: config/diet/default.yaml | Schema: (none) Scoring weights (dietary 30%, consistency 25%, check-in 25%, device 20%), lookback window, trend calculation, LLM model for diet insights, and score thresholds (good >= 7, warning >= 4). Access: get_diet_config()

​

demo — Demo/Seed Data

File: config/demo/default.yaml | Schema: (none) Demo Clerk user ID for seed data. Not auto-composed — must be explicitly included.

---

​

Adding New Configuration

1. Create config/<group>/default.yaml with your fields
2. Add - <group> to config/default.yaml defaults list
3. (Optional) Create schemas/<group>.py with a typed dataclass
4. Add get_<group>_config() accessor to hydra_config.py
5. Use env overrides: field: ${oc.env:MY_VAR,default_value}