Changelog

1.1.0

SQLite Database Provider

• Added SQLite as the default database backend — zero external dependencies
• Full multi-tenant support with per-tenant .sqlite files
• All 16 domain mixins implemented (tenants, users, projects, models, vectors, tracing, guardrails, RAG, memory, alerts, prompts, quotas, files, providers, inference, API tokens)
• WAL mode and foreign keys enabled for performance and data integrity
• Switch to MongoDB with DB_PROVIDER=mongodb

SQLite Vector Provider

• Added sqlite-vector provider contract for local vector storage
• Brute-force cosine, dot-product, and euclidean similarity search
• No external vector database required
• Batch upsert within transactions for data consistency

Local File Provider

• Already available as local-file provider

Quick Installation

• Zero-dependency setup: just npm install && npm run dev
• Docker runs with a single volume mount for persistence
• Updated documentation with Quick Installation section

Configuration

• New env vars: DB_PROVIDER (default: sqlite), SQLITE_DATA_DIR (default: ./data)
• MONGODB_URI now only required when DB_PROVIDER=mongodb

---

1.0.0

Initial release of the Cognipeer Console documentation.

Core Infrastructure

• Centralized configuration via getConfig()
• Winston structured logging with request context
• AsyncLocalStorage-based request context propagation
• Provider-based cache (none / memory / Redis)
• Retry with exponential backoff + circuit breaker
• LRU runtime pool for provider SDK instances
• Fire-and-forget async task runner
• Health check registry
• Graceful lifecycle management
• Configurable CORS for client APIs

Features

• Multi-tenant architecture with per-tenant databases
• JWT session + API token authentication
• Contract-driven provider system (LLM, Vector, File)
• OpenAI-compatible chat completions and embeddings
• Vector store management (multi-provider)
• Agent tracing (batch + streaming)
• Guardrails (PII, moderation, prompt shield, custom)
• RAG pipeline (ingest, chunk, embed, query)
• Versioned prompt templates with environment deployment
• Semantic memory stores
• File storage with Markdown conversion
• License-based feature control
• Quota and rate limiting

Deployment

• Docker multi-stage build
• Kubernetes Helm chart
• GitHub Pages documentation