Download

For training instructors

Train faculty on agents.

Run workshops on modern AI workflows, agentic coding included, with a local model on your own hardware. Admins decide which skills and MCP servers are on, so the training environment is one you control.

See how it worksDownloadTalk to us
The Agent Skills settings: skills an admin switches on or off, such as document creation, charts and a SQL connector, plus custom skills, agent flows and MCP servers.

Runs with

  • Ollama
  • LM Studio
  • LocalAI
  • KoboldCPP
  • Docker
  • Kubernetes
  • Helm
  • OpenShift

What it gives you

A training lab you control.

The Agent Skills settings: skills an admin switches on or off, such as document creation, charts and a SQL connector, plus custom skills, agent flows and MCP servers.

Switch built-in skills on or off, add custom skills, chain steps in the no-code flow builder, and connect MCP servers over stdio, SSE or streamable HTTP.

See the agent controls

Show how agents plan, call tools and work with files. The filesystem skill (Docker installs) works in a folder under the server's storage.

Ollama, LM Studio, LocalAI, KoboldCPP, NVIDIA NIM or any OpenAI-compatible server, on hardware you run.

Give each workshop workspace, and its agents, their own provider and model.

Invite each instructor, then review the workshop chats and the admin event log.

How it works

Run a workshop in three steps.

  1. Turn on the skills

    Switch on the built-in skills the workshop needs: document search, summaries, charts, file creation and SQL.

  2. Connect MCP servers

    Add Model Context Protocol servers to give agents the tools your workshop teaches.

  3. Hand a task to an agent

    Type @agent, or use Automatic mode with a model that supports tool calling, and follow each tool call in the chat.

Before you rely on it

Good to know.

Plain facts about what this use can and cannot do today.

See the agent controls

  • Mission LLM is not a code editor. Agentic coding is what the training covers, run on the app's agents.
  • Web search and cloud models connect out only when you configure them.
  • Agents work best with a model that supports tool calling.

More ways to use it

  • Classrooms

    A course assistant that answers from the course readings and lists its sources.

  • Students

    Study help drawn from the assigned material, in a workspace of your own.

  • Teaching a course

    Build a workspace per course from the syllabus, readings and handouts.

  • Teams and labs

    Policies, manuals and research, searchable with the sources listed.