Analysis updated 2026-05-18
Stand up a Slack support bot that answers questions in a channel, running in its own hardened container.
Track exactly how many tokens and how much cost each agent conversation uses via Langfuse.
Deploy the same agent configuration locally with Docker Compose or to production on AWS or GCP Kubernetes.
Give an agent persistent memory and a scheduled cron job without writing custom infrastructure code.
| xatuke/wrapper | 4ssh1/mine-sweeper | achawla19/intuitcode-extension | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | TypeScript | TypeScript | TypeScript |
| Last pushed | — | 2026-01-31 | — |
| Maintenance | — | Maintained | — |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 4/5 | 2/5 | 4/5 |
| Audience | developer | vibe coder | developer |
Figures from each repo's GitHub metadata at analysis time.
Needs Node.js, Docker, a Slack workspace, and an Anthropic API key to get a first agent running.
wrapper is a command line tool for running AI agents on Slack, where each agent lives inside its own isolated container and every conversation is traced for cost, tokens, and tool usage using an open source tracing tool called Langfuse. The project treats an agent as nothing more than configuration plus tools, so adding a new agent means writing a small YAML file and optionally a folder of custom tool code, rather than building a new service from scratch. Each agent gets its own Slack app that connects over an outbound only connection, meaning nothing needs to be exposed to the public internet, along with its own storage for conversation history and per message traces. Optional features let an agent schedule its own recurring tasks, keep persistent memory between conversations, and work with files in its own workspace. Custom Docker images can add extra command line tools or packages, and longer secrets can be stored safely in an encoded form. Getting started involves installing dependencies, copying an example configuration file, running an init command that starts the tracing stack and sets up the organization, adding an API key, then creating and starting an agent with a couple of commands before inviting it into a Slack channel. From there, a small set of day to day commands cover checking status, viewing logs, updating environment variables, restarting, and removing an agent. The tool supports two ways of running: locally with Docker Compose on your own machine or a cloud server, or on Kubernetes for AWS or Google Cloud, using the exact same commands either way. Files that Kubernetes or Docker Compose need are generated automatically by the tool and are not meant to be edited directly, while each user's own agent configurations live in their own separate, git ignored folder so updates to the framework do not touch a user's custom agents.
A CLI framework for running config-defined AI agents on Slack, each in its own container with built-in cost and usage tracing.
Mainly TypeScript. The stack also includes TypeScript, Docker, Kubernetes.
Setup difficulty is rated moderate, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.