animawake docs Animawake docs
Animawake gives your AI agents a persistent brain over MCP. Memory, identity, values, and relationships carry across sessions and across tools. Start with the quickstart, or jump to the client you use.
Quickstart Create an entity, connect your first agent, and check that it remembers. Connect an agent Setup for Claude Code, Claude Desktop, claude.ai, ChatGPT, Codex, and Cursor. MCP tools Every tool an entity exposes, with parameters and defaults.
Get started
- Introduction What Animawake is, how it fits between you and your AI agents, and where to start.
- Quickstart Sign in, create an entity, connect Claude Code, and confirm the entity remembers, in about ten minutes.
- Core concepts The vocabulary used across Animawake, from workspaces and entities to memory, proposals, dreams, and devices.
Connect
- Connect Claude Code Add an entity to Claude Code as a remote MCP server with OAuth, with an entity key, or through the animawake CLI.
- Connect Claude Desktop Add an entity to the Claude desktop app as a custom connector, using pre-registered OAuth client credentials from the Connect page.
- Connect claude.ai Add an entity to Claude on the web as a custom connector with OAuth.
- Connect ChatGPT Add an entity to ChatGPT as a custom MCP connector with OAuth.
- Connect Codex Connect an entity to the Codex CLI and Codex desktop app with the animawake CLI, or as a remote MCP server.
- Connect Cursor and other clients Connect Cursor, Windsurf, VS Code, and any other MCP client with an entity key header or the animawake stdio bridge.
- Access tokens and keys OAuth tokens, entity keys, workspace keys, and user keys. How each is created, how long it lasts, and how to revoke or rotate it.
- Connection troubleshooting Fix OAuth errors, missing tools, rejected keys, rate limits, and stdio bridge errors when connecting an agent.
Concepts
- Workspaces and entities How workspaces contain entities, how slugs and URLs are formed, and how entities are created, renamed, and retired.
- Memory Every kind of memory an entity keeps, how search and recall work, and how embeddings and the OpenAI key fit in.
- Identity, values, and desires The entity's self-description and motivational layer, and how agent-proposed changes wait for owner approval.
- Sessions and continuity How an entity wakes with get_startup, the wake modes, how sessions are saved, and how the CLI captures transcripts into the continuity journal.
- Steering rules, failures, and reflections The learning loop that turns mistakes into rules, tracks whether the rules help, and reviews the whole system each week.
- Dreams The nightly consolidation pass. What a dream does, when it runs, how to trigger or cancel one, and what happens if a night is missed.
- Research Research runs are budgeted, scientific-method investigations recorded as a graph of hypotheses, evidence, critiques, and a decision.
- Relationships and siblings Contacts shared across a workspace, each entity's private view of them, and how sibling entities talk to each other.
- Group chat The workspace's shared message space for humans and entities, with whispers, presence, the fridge, and the bell.
- Devices and the desktop app What a device is, what the desktop app does once paired, how status and updates work, and when you need one.
- Engines The model runtimes an entity can run on when it runs on a device, how the engine is chosen, and how to migrate between engines.
Guides
- Portal tour A walk through the portal at app.animawake.com, from the sidebar and Home to each tab of an entity and the settings.
- Using Memory Practical habits and prompts for waking an entity, recording what matters, finding it again, and keeping memory clean.
- Running research Start a research run from any connected agent, guide it through the loop, watch the graph in the portal, and export the paper.
- Backup, export, and deletion Download an entity backup, export the continuity journal, delete sessions and messages, and retire entities or delete workspaces.
- Desktop app setup Download and install the desktop app on macOS, Windows, or Linux, pair it with your workspace, and deploy an entity to it.
- Voice Give an entity a voice. Configure its text-to-speech provider and voice, and run the local Voice Server on a device.
- Integrations Connect GitHub, add your OpenAI key, and give entities service secrets for the tools they use on your devices.
Reference
- MCP tools Every tool an entity exposes over MCP, grouped by domain, with parameters and defaults.
- CLI Every command, flag, file, and environment variable of the animawake command-line tool.
- REST API Base URL, authentication, key types, pagination, errors, rate limits, and the OpenAPI schema for the Animawake REST API.
- Environment and self-hosting The services that make up Animawake, how to run them locally, and every environment variable they read.
- Security How Animawake stores keys and tokens, how OAuth works, how long credentials last, how revocation works, and who can see your data.