Packages

54 packages, all MIT, all alpha (0.1.x). The two foundation packages — oridecon and oridecon-contracts — are all you need to start. Everything else is an extension: install only what the app uses.

Every extension depends only on oridecon and oridecon-contracts, never on another extension.

oridecon-contracts   →   Zero dependencies (protocols, types, exceptions)
oridecon             →   Application, container, providers, modules, Result, YAML config
oridecon-*           →   Extensions (web, SQL, auth, AI, …)

Start from the CLI. Templates pick the first handful of packages; you add the rest later.

uv add oridecon-cli
oridecon new project my-app --template web-api
oridecon run

This catalog is the same list as The Ecosystem and the sidebar. Canonical URLs live under /packages/{group}/…, /platform/…, and /experimental/…. Ungrouped /packages/oridecon-* paths redirect here.

Foundation

Application, container, providers, modules, config, Result.

Package	What it does
`oridecon`	Async-first DI container, Application lifecycle, modules, providers, config, and the Result type.
`oridecon-contracts`	Zero-dependency protocols, value types, and exceptions shared across every package.

Web & HTTP

ASGI, GraphQL, outbound HTTP, webhooks.

Package	What it does
`oridecon-web`	ASGI web layer — controllers, routing, middleware, OpenAPI, CORS, CSRF, rate limiting.
`oridecon-http`	Outbound HTTP client with resilience and observability built in.
`oridecon-graphql`	GraphQL server (Strawberry) — schema, subscriptions, depth/complexity limits, persisted queries.
`oridecon-webhook`	Webhook management — subscriptions, delivery tracking, HMAC verification, dead-letter queue.

Data

SQL, NoSQL, cache, object storage, search.

Package	What it does
`oridecon-sql`	Async SQL for Postgres / MySQL / SQLite — repositories, migrations, query building.
`oridecon-nosql`	Document-store abstraction (MongoDB, DynamoDB, Firestore).
`oridecon-cache`	Multi-backend caching — Redis, Memcached, in-memory — behind CacheBackendProtocol.
`oridecon-storage`	Unified blob storage — S3, GCS, Azure Blob, R2, local filesystem.
`oridecon-search`	Full-text search — Elasticsearch, Meilisearch, Typesense, and SQL backends.

Security

Auth, secrets, JWT, OAuth2, guards.

Package	What it does
`oridecon-auth`	Authentication and authorization — JWT, OAuth2, RBAC, password hashing, web guards.
`oridecon-secrets`	Secret vaults with rotation, tenant scoping, and audit.

Events

Domain events, queues, notifications.

Package	What it does
`oridecon-events`	Event sourcing and CQRS — domain events, aggregates, command/query buses, projections.
`oridecon-queue`	Message bus / queue with named multi-backend support (Redis, RabbitMQ, Kafka, SQS).
`oridecon-notification`	Email, SMS, and push delivery with multi-backend support.

Infrastructure

Tasks, workflows, tenancy, resilience, vectors.

Package	What it does
`oridecon-tasks`	Background tasks — scheduling, workers, and job queues over memory / Redis / AMQP / Postgres.
`oridecon-workflow`	Workflow orchestration — pipelines, bulk ops, sagas, graph engine.
`oridecon-tenancy`	Multi-tenant resolution, lifecycle, and isolation strategies.
`oridecon-resilience`	Circuit breaker, retry, bulkhead, rate limiting, throttle, fallback.
`oridecon-monitor`	Health checks, metrics, tracing, and structured logging (Prometheus / OpenTelemetry).
`oridecon-audit`	Append-only, HMAC-verified, retention-managed audit trail.
`oridecon-vector`	Vector store backends (pgvector, Qdrant, Pinecone, in-memory) for embeddings.
`oridecon-graph`	Graph database support (Neo4j, in-memory).
`oridecon-features`	Feature-flag management with caching and pluggable providers.

AI platform

LLMs, RAG, agents, memory, MCP, skills, eval.

Package	What it does
`oridecon-ai`	Orchestration layer that discovers and wires the AI subsystems below.
`oridecon-ai-llm`	Multi-provider LLM client with routing and thinking suppression.
`oridecon-ai-rag`	Retrieval-augmented generation — chunking, retrieval, synthesis, citations.
`oridecon-ai-agents`	Agents with tools and strategies (ReAct, plan-and-execute).
`oridecon-ai-memory`	Episodic, semantic, and working memory for AI systems.
`oridecon-ai-skills`	Skill/tool registry, executor, and built-in tools.
`oridecon-ai-session`	AI conversation sessions — branching, checkpointing, multi-agent.
`oridecon-ai-mcp`	Model Context Protocol server and client for AI agents.
`oridecon-ai-workers`	Background AI work — batch embedding, document ingestion, maintenance.
`oridecon-ai-feedback`	Collect, process, and store feedback on AI responses.
`oridecon-ai-guard`	Input/output guard pipeline — LLM safety and content filtering.
`oridecon-ai-governance`	AI governance — policy enforcement, audit trails, budget tracking.
`oridecon-ai-evaluation`	Evaluation framework — benchmarks and quality gates for AI outputs.
`oridecon-ai-prompt`	Prompt management — templates, composition, optimization.
`oridecon-ai-observability`	Tracing, metrics, and health checks specific to AI calls.
`oridecon-ai-relay`	AI relay — route and fan-out model calls across providers.
`oridecon-ai-relay-gateway`	Gateway for the AI relay — ingress, auth, and quota at the edge.

Multimedia

Image, video, audio, TTS, upscale, interpolate.

Package	What it does
`oridecon-multimedia`	Orchestration layer for multimedia generation packages.
`oridecon-multimedia-image`	Image generation backends.
`oridecon-multimedia-video`	Video generation backends.
`oridecon-multimedia-tts`	Text-to-speech backends.
`oridecon-multimedia-music`	Music generation backends.
`oridecon-multimedia-beat`	Beat / rhythm generation backends.
`oridecon-multimedia-upscale`	Image and video upscaling backends.
`oridecon-multimedia-interpolate`	Frame interpolation backends.

Tooling

CLI, testing, admin, UI — experimental except testing.

Package	What it does
`oridecon-testing`	Fakes, test clients/beds, fixed clock, and protocol compliance suites.
`oridecon-cli`	oridecon command — project scaffolding, dev server, migrations, inspection.
`oridecon-admin`	Admin panel — CRUD, dashboards, RBAC.
`oridecon-ui`	HTMX/htpy component library.

Next Steps

The Ecosystem — grouped by purpose, with install examples
Adoption Paths — starter → standard → advanced → AI
Architecture — why extensions never import each other
Installation — pin 0.1.x and scaffold