Which Snipara Package Should I Use?

Most users only need one of two entry points: snipara-mcp for MCP access, or create-snipara when they also want rlm-runtime.

Default Recommendation

Use snipara-mcp for MCP tools, OAuth login, project-scoped context, and memory.
Use create-snipara when you also want rlm-runtime.
Only install specialized packages when you already know you need them.

Quick Matrix

Package	Use it for	Audience	Status
`snipara-mcp`	MCP access, OAuth login, context queries, memory, stdio MCP server	Most users	Core
`create-snipara`	One-command onboarding for Snipara MCP + RLM Runtime	Most users who want setup automation	Core
`snipara`	Python SDK, programmatic workflows, config-driven Python integration	Python developers	Active
`snipara-memory`	Standalone memory engine when memory itself is the product you want	Specialized teams and OSS adopters	Specialized
`snipara-orchestrator`	Production-first validation, proofs, drift detection, advanced agent workflows	Advanced teams	Advanced
`rlm-runtime`	Sandboxed code execution runtime that complements Snipara	Users who need execution, not just context	Adjacent
`snipara-openclaw-hooks`	OpenClaw-specific hook automation	OpenClaw users	Specialized
`snipara-companion`	Optional local helper CLI and hook-oriented workflows	Advanced local automation users	Optional
`snipara-openclaw-install`	Legacy OpenClaw onboarding wrapper	None for new installs	Archived

Choose by Use Case

MCP for Codex, Claude Code, Cursor, ChatGPT, Gemini, VS Code

Start with snipara-mcp. If you also want local code execution via rlm-runtime, use create-snipara.

Codex specifically

Use mcp.json plus AGENTS.md. You do not need snipara-companion unless you have a very specific local helper workflow.

Python application integration

Use snipara when you want a Python SDK. Use snipara-mcp when the goal is simply MCP client setup.

Production validation and orchestration

Use snipara-orchestrator only if you need proof-based validation and production drift detection. It is an advanced package, not the default install path.

Companion vs orchestrator

Do not merge snipara-companion and snipara-orchestrator. They solve different problems: companion is a local helper CLI for hook-oriented workflows, while orchestrator is an advanced Python package for production validation and orchestration.

OpenClaw

Use create-snipara for base Snipara setup, then add snipara-openclaw-hooks for OpenClaw-specific automation. The older snipara-openclaw-install wrapper is archived.

Recommended Paths

Fastest General Setup

npx create-snipara

Best choice when you want setup automation and may also need RLM Runtime.

Core MCP Only

pip install snipara-mcp

Smallest, clearest install when all you need is Snipara MCP.