Analysis updated 2026-05-18
Chat with your local pi coding agent remotely through Telegram or Discord.
Manage which chat users or groups can reach your agent with deny-by-default pairing.
Run multiple project workspaces with different tool permissions and model choices.
| ceifa/eleven | 0labs-in/vision-link | alfons-fhl/cursor-plan2api | |
|---|---|---|---|
| Stars | 4 | 4 | 4 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 3/5 | 3/5 |
| Audience | vibe coder | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires Node.js 24+ and existing pi credentials before the dashboard can connect a workspace.
eleven is a lightweight gateway that lets you talk to an AI coding agent called pi from chat apps like Telegram and Discord, instead of only from a terminal on your own computer. The idea is that pi is an agent you already run locally, in a specific project folder, and eleven simply makes that same agent reachable from your phone or a group chat when you are away from your machine. It was built as a smaller, simpler alternative to a similar project called OpenClaw. eleven itself stays deliberately minimal. It has no memory system, no scheduler, and no plugin system of its own, only around 3,000 lines of code on top of the underlying pi toolkit, which handles the actual agent behavior, more than 30 model providers with sign-in support, tool use, and saving conversation sessions. Because everything about how the agent behaves lives in the project folder itself, your existing configuration files and skills continue to work unchanged whether you talk to the agent locally or through eleven. The tool is organized around a few core concepts. Channels connect the agent to chat platforms like Telegram or Discord. Pairing controls access by denying unknown senders and unregistered groups by default, showing them as pending requests you approve with one click in a web dashboard. Workspaces define a project folder along with what the agent is allowed to do there, ranging from read-only access to full command execution, plus an optional choice of AI model. The dashboard itself runs locally on your machine and lets you configure channels, approve access requests, and watch or join live conversations as they happen. Model provider settings support an ordered list of fallback models, so if one model errors or returns an empty response, the request automatically retries with the next one. Installing eleven requires Node.js version 24 or newer and existing credentials for pi. Once installed, a single dashboard page is used to add a chat bot token and connect a workspace, and the project also offers a command to import existing settings from OpenClaw.
A lightweight gateway that connects the pi AI agent to chat apps like Telegram and Discord, with a local web dashboard.
Mainly TypeScript. The stack also includes TypeScript, Node.js, pi SDK.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Verify against the repo before relying on details.