Deploy LibreChat as a self-hosted agent workspace
Deploy LibreChat when an operator needs a self-hosted, multi-user agent workspace with model routing, MCP tools, skills, code execution, access controls, and audit-friendly administration.
npx skills add agentskillexchange/skills --skill deploy-librechat-as-a-self-hosted-agent-workspace
Use LibreChat when an operator needs to run a self-hosted agent workspace instead of sending users to separate hosted chat products. The workflow is to deploy LibreChat, configure model providers and endpoints, enable Agents with MCP tools and reusable Skills, set role and group permissions, and expose a controlled workspace where users can run file search, code execution, artifacts, web search, and custom assistants.
What this skill actually does
When to use it
Invoke this skill when a team needs a private or organization-managed AI workspace with repeatable agent configuration, provider routing, and administrator controls. It is especially useful when the agent environment needs to support multiple model providers, MCP servers, custom endpoints, shared agents, code interpreter workflows, and observability without relying on a single hosted chat UI.
Scope boundary
This is not a generic LibreChat product listing. Keep the workflow bounded to deployment and operation of LibreChat as an agent runtime surface: configure providers, agents, tools, skills, permissions, and operational controls. Do not use it for ordinary one-off chat usage or broad application development unrelated to LibreChat administration.
Operator checklist
- Choose a supported deployment path and configure storage, authentication, and administrator access.
- Add the required model providers or OpenAI-compatible endpoints.
- Configure LibreChat Agents, MCP servers, reusable Skills, file search, code execution, and tool permissions.
- Set role and group controls before sharing agents with users.
- Verify observability, upload limits, workspace access, and security settings before production use.