OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Infrastructure

Mirage

Mirage is open-source infrastructure that provides AI agents with a virtual terminal joining a virtual filesystem, virtualized command-line tools and configurable runtimes. Developers can mount data sources under paths, route commands to runtimes and govern agent actions and filesystem visibility through profiles and policies.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Mirage — OpenViST project profile in Agent Infrastructure.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Mirage when building an agent application that benefits from one shell interface across mounted data, service CLIs and routed runtimes, particularly when command permissions and filesystem visibility need to be configured.

Who it is for

  • Developers integrating AI agents with multiple data sources through a shared filesystem and shell.
  • Teams that need to route agent commands to configurable runtimes.
  • Agent builders who want controls over commands, tools and the data an agent can see.

Documented uses

  • Mount data sources side by side and search or copy files across them using shell commands.
  • Expose service operations through virtualized CLIs with separately configured names and credentials.
  • Run scripts through configured runtimes, including sandboxed Python in the documented example.
  • Set command and CLI permissions and control which files or directories are visible to an agent.
  • Snapshot a workspace in the documented quickstart.
Before you choose

Requirements & limits

The repository labels the project preview and identifies it as Apache-2.0 licensed. The README lists Python 3.11 or later for the Python package and Node.js 20 or later for the TypeScript SDK.

  • Mirage is an integration layer for agent applications; the examples create workspaces from application code or the CLI.
  • The Python and TypeScript paths have documented minimum runtime versions.
  • The virtual filesystem, command-line interfaces and runtimes are joined by one shell, allowing commands and pipelines to span those components.
  • Profiles support allow, ask and deny for commands and CLIs, plus hide and show for files and folders; a scriptable policy engine can prohibit actions before they run.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a Mirage workspace with Python

Install the Python package and define a workspace with RAM and S3 mounts.

Before you start

  • Python 3.11 or later
  • An S3 bucket configured for the example

Steps

  1. Install the package with the command below.

    Documented command
    uv add mirage-ai

  2. Import Workspace, RAMVFS, S3Config and S3VFS from the documented Mirage modules.

  3. Create a workspace that mounts RAMVFS() at /data and an S3VFS configured with the bucket at /s3.

Expected outcome

A Python workspace definition with RAM and S3 filesystems mounted at the documented paths.

Guide limitations

  • The setup example does not document S3 credential configuration.
  • This guide covers installation and workspace definition, not execution or deployment.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
10 October 2026
Original website title
strukto-ai/mirage · Official homepage title · retained in its source language
How we review sourcesSuggest a correction