TeamCopilot is an open-source AI agent that lets teams create and reuse AI skills and workflows from a shared web interface. Instead of every teammate cloning repos, configuring dev environments, managing API keys, or learning how to run internal scripts, engineers can set up reusable workflows once and make them available to the rest of the team through chat. It supports fine-grained permissions, scoped secrets, approvals, scheduled runs, audit logs, self-hosting, and multiple LLM providers.
Hey Product Hunt 👋
I’m Rishabh, and I’m building TeamCopilot.
The idea came from a problem inside my own org: non-technical teammates wanted to use AI agents like Claude Code for work, but setup was too hard.
We had custom Claude skills, tools, scripts, API keys, SSH keys, and dev-environment requirements. Engineers could handle that. Non-technical teammates did not want to clone repos, configure environments, manage secrets, or keep local setups updated just to use an AI agent.
So I built TeamCopilot. A few things it supports today:
- Reusable AI skills and workflows
- No local setup for end users
- Fine-grained permissions for skills, workflows, and tools
- Scoped secrets for shared and personal credentials
- Approval flow before workflows are made available
- User approval before sensitive workflows run
- Scheduled agent sessions and workflow runs
- Audit logs for chats and workflow executions
- Self-hosting, so code and data stay on your infra
- Model-agnostic LLM support
It’s open-source and still early, so I’d love feedback from teams experimenting with internal AI agents. Would love your thoughts, feedback, and criticism.
Report
No reviews yetBe the first to leave a review for TeamCopilot
SuperTokens