The World's First Virtual Terminal for AI Agents

Agents don't need a sandbox, they need the terminal they were trained on.

npm install @struktoai/mirage-node
npm install @struktoai/mirage-browser

const ws = new Workspace({
  '/tmp':   [new RAMVFS(), MountMode.EXEC],
  '/redis': [new RedisVFS({ url: redisUrl }), MountMode.WRITE],
  '/slack': [new SlackVFS({ token: slackToken }), MountMode.EXEC],
}, {
  runtimes: [buildRuntime('monty', { captures: ['python', 'python3'] }), 'workspace'],
})
// one grep sweeps every source
await ws.shell('grep -rln session /redis /tmp')
// run a script that lives in Slack, report into Redis
await ws.shell('python3 /slack/report.py > /redis/report.txt')
// read the report back with a POSIX-style call
const report = await ws.vfs.read('/redis/report.txt')
// a typed CLI under a head word, found by man and which
ws.registerCli('slack', SLACK, { token: slackToken })
await ws.shell('slack send-message --channel general --text "report is up"')

Virtual Terminal Overview

A Unified VFS
Virtual CLIs
The Mirage Shell
Virtual Runtimes
Policy Engine
Auth and Notifications
Profiles and Agents
Portability

A Real Example

A real Mirage workspace, running in your browser.

1mounts:
2  /slack:
3    vfs: slack?Your Slack, read over its API and mounted as folders: one per channel, one per day, the day's messages as chat.jsonl.
4    config:
5      token: 
6      search_token: 
7      content_search: true?Lets grep -rw ask the service's own search index which files can match, and read only those, instead of every file under the mount.
8    mode: read