Eidola MCP Server
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Eidola MCP Serverlist available engrams"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Eidola Runtime
AI Personality Runtime — portable persona system with reactive avatars.
Eidola gives AI personas a visible, reactive presence, inside the editors you already use. Install one CLI, awaken a persona ("Engram"), and a local display (Shrine) shows an animated avatar (Vessel) that reacts to what your agent is doing.
This repo is the open, MIT-licensed runtime — the local-first middleware you install and run yourself. It's not a chat app; it sits on top of your editor via MCP and hooks. No lock-in, fully auditable.
Browse, create, and publish Engrams at Eidola.app.
Install
1. Install the runtime
npm install -g "@eidola/cli"Requires Node.js 20 or later.
2. Create your Eidola folder
Create a folder called Eidola anywhere on your computer (e.g.
~/Documents/Eidola/) and download an Engram into it — unzip it directly
inside.
3. Connect your AI editor
Run one of these once, then fully quit and relaunch your editor:
eidola setup-claude # Claude Code
eidola setup-cursor # CursorEach writes the Eidola MCP server entry plus the hook relay config that
drives a reactive Vessel from tool activity (workspace-scoped by default;
pass --global to install to your home directory instead).
4. Launch the Shrine
eidola launch shrineOpen the shrine in your browser at http://127.0.0.1:9743/shrine/. Or just ask your
editor's agent to "launch the shrine."
5. Awaken
In the Shrine UI: choose your Eidola folder → pick an Engram → click Awaken. Optionally press F11 for fullscreen.
You can do the same thing from chat instead — ask your agent to "Awaken
<engram-id>" — it's equivalent to clicking Awaken in the UI.
Related MCP server: solitaire-ai
CLI commands
Command | What it does |
| Starts the MCP server over stdio. Your editor's MCP config invokes this — you won't normally run it by hand. |
| Starts the Shrine display ( |
| Stops a running Shrine display. |
| Adds the Eidola MCP server to Cursor and installs the reactive-Vessel hooks. |
| Adds the Eidola MCP server to Claude Code and installs the reactive-Vessel hooks. |
| Removes the Eidola MCP server entry and hooks from Cursor and Claude Code, plus any |
MCP tools
For driving Eidola from chat instead of the Shrine UI:
Tool | Input | What it does |
|
| Loads an Engram, binds its Vessel, wires up Soul delivery for your editor, and shows it on the Shrine display. |
| (none) | Puts the active Engram to sleep — removes the Soul artifacts |
|
| Starts the Shrine display if it isn't already running. |
See docs/spec/cli-mcp-reference.md for full details.
What's an Engram
An Engram is a portable persona package:
{id}/
SOUL.md # personality, written in prose
vessel.yaml # expression map (state → clip)
engram.yaml # metadata/version
vessels/{pack}/
idle.mp4 # or .gif / .json / .webm
thinking.mp4
...The full spec lives in docs/spec/engram-format.md.
What's in this repo
Package | Published as | What it does |
| MCP server + CLI. Runs over stdio ( | |
|
| Shared tool-name → Vessel visual-state mapping, used by both hook relay packages and the Cursor extension. |
|
| Reads Cursor agent hook events and relays them to the local state socket. |
|
| Reads Claude Code hook events and relays them to the local state socket. |
|
| The Shrine — a local HTTP server that renders the active Vessel (Lottie/WebM/Three.js) and reacts to live state over SSE. |
|
| Optional VS Code extension for coarser LM-state signals inside Cursor. Not published under the |
Building from source
pnpm install
pnpm build
pnpm testEach package can also be built/tested individually with pnpm --filter <package> run build.
Contributing
This repo is young and still stabilizing its public release shape. Issues
and PRs are welcome — please open an issue to discuss substantial changes
before sending a PR, since the Engram spec format and the hook
relay/state-socket protocol are both meant to stay stable across installs.
Run pnpm build && pnpm test before submitting.
License
MIT — see LICENSE. The Engram spec format is, and will remain, open regardless of what gets built or monetized on top of it.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- Alicense-qualityCmaintenanceIdentity infrastructure for AI agents. Gives agents an evolving persona, session continuity, and self-correcting retrieval so they stop being strangers. Local-first, model-agnostic.Last updated8AGPL 3.0
- Alicense-qualityBmaintenanceGive your AI coding assistant live visibility into the user's browser. Claude sees the actual DOM, console errors, and network timing instead of guessing from screenshots. Drop-in middleware for FastAPI + Flask.Last updatedMIT
- Alicense-qualityDmaintenanceThe simplest way to bridge and collaborate across AI Agent sessions like Claude Code, Codex, Gemini, or Cursor. It allows your agents to combine their strengths to solve your most difficult tasks without leaving their current context.Last updated4362MIT
Related MCP Connectors
The team layer for AI coding agents: shared contracts, collision alerts, E2EE sessions.
Cross-agent artifact workspace with provenance across Claude Code, Codex, Cursor, LangGraph.
User-owned memory for AI agents, Copilot, Claude, IDEs, CLIs, and chat apps over remote MCP.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/AyeJK/eidola-runtime'
If you have feedback or need assistance with the MCP directory API, please join our Discord server