Développement
open-terminal
Open Terminal — a computer you can curl: a sandboxed terminal API for AI agents, from Open WebUI.
★ 3,3 k étoilesMIT
Voir le guide →mirage est un projet de développement open source, créé par strukto-ai (strukto-ai/mirage sur GitHub). Il compte 3 681 étoiles sur GitHub et 272 forks, il est publié sous la licence Apache-2.0 et a été mis à jour pour la dernière fois sur GitHub le 8 octobre 2026.
A virtual terminal and filesystem that lets AI agents work across S3, Drive, Slack, Gmail and more.
Decide the first concrete outcome you want from it, so you can pick the right example, template or mode instead of exploring everything at once.
Clone the repository (or install the published package) and skim its README quick start for the requirements for your platform.
Run the documented install step and set any required configuration — such as provider keys, data sources or local services — before the first run.
Execute the smallest documented example end to end, confirm the output is what you expected, then adapt it to your real data or workflow.
Many sources mounted side by side under one root, reachable with standard unix tools.
Tools such as git and slack are answered by Mirage itself — nothing to install per service, with separate credentials per virtual tool.
Python, JavaScript and other commands run in-process, sandboxed or remotely, decoupled from storage.
allow / ask / deny govern commands and CLIs; hide / show govern files, so hidden paths are absent, not just unreadable.
A scriptable policy can block dangerous actions before they run, gating every VFS operation and session write.
External changes arrive as an event stream on the mount, so agents react instead of rescanning.
Get the source so you can inspect, configure and run it locally.
git clone https://github.com/strukto-ai/mirage.git
cd mirageInstall the Python package/CLI or the TypeScript SDK as the README's Installation section describes.
uv add mirage-ai # installs the mirage library and the mirage CLI binary
# or run without installing: uvx mirage-aiFollow the README quickstart (Python, TypeScript or CLI) to mount a source and run a first command across it.
mirage # see README quickstart for the exact first commandsIt is maintained by strukto-ai on GitHub (strukto-ai/mirage), where the source, issues and releases live.
Yes — it is free and open source under Apache-2.0, which permits use, modification and commercial redistribution, with licence/notice retention.
Mirage itself needs no LLM key; you connect your own service credentials, and your agent brings its own model access.
Compared with writing separate SDK integrations or MCP servers per service, mirage's distinguishing focus is: a virtual terminal and filesystem that lets AI agents work across S3, Drive, Slack, Gmail and more. Pick it when that focus matches your need; the alternative may suit you better if its broader or different scope is the priority.
Clone https://github.com/strukto-ai/mirage (or install its published package), follow the README quick start to install and configure it, then run the smallest documented example — the Run section on this page mirrors those steps.
Open Terminal — a computer you can curl: a sandboxed terminal API for AI agents, from Open WebUI.
agtx: a terminal Kanban blackboard where Claude, Codex, Grok, and more collaborate on the same tasks.
Turn any website into a CLI — and let AI agents drive your already-logged-in browser
Context optimization for coding agents: sandbox tool output, keep sessions lean, remember more
One Rust CLI for every LLM: shell assistant, chat REPL, RAG, tools and agents across 20+ providers.
Gmail for coding agents: identities, inboxes, threads, and file leases over FastMCP, Git, and SQLite.
Les autres projets de la catégorie Développement avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si mirage ne vous convient pas.
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
Projets de la catégorie Développement proches dans le catalogue, les plus proches en étoiles de mirage — continuez à explorer de part et d'autre.
Like htop, but for AI coding agents — watch Claude Code and Codex sessions in real time.
Open-source AI coding assistant for VS Code. Code completion, chat, edits and reviews with local or hosted models. Your models, your infrastructure.
The official Java SDK for building Model Context Protocol servers and clients.
A streamlined framework for evaluating and benchmarking LLMs, VLMs and AIGC models.
The agent harness for complex codebases inside Codex — project memory, planning and verified completion.
A Node.js crawler that turns websites into LLM-ready data and extracts structured SERP results.
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.