Hey everyone 👋

This is a problem we keep hitting: every team building handles the same plumbing on top of using cloud code, codecs, or any agent. Spin the sandbox, clone the repo, install the deps, wire the MCPs, expose the chat endpoint, tunnel the dev server.

AnyFrame is the bottom layer. You specify the agent once (repo, install cmd, skills, MCPs) and it creates a cached sandbox image. Boot a session and chat with it from the web UI or Python SDK. Plug in the MCP connectors (Linear, Sentry, Google, …) once and toggle them per agent.

Some things you can make above:
– An internal “chat-deployed” agent for your team
– A scoped coding agent for your customers, locked in their repo.
– A per PR review bot that actually runs your code, not just reads it.
– Honestly anything you’d otherwise be a hand-wiring sandbox for.

Free to try – bring a Claude OAuth token or Codex API key and you’re in.

🌐 Product: https://anyfrm.com
🐍 Python SDK + Demo: https://github.com/tinyhq/anyfra…
💬 Differences: https://discord.gg/UpkEW6JjpU

We’ve been building AnyFrame for the past couple of weeks and we’re excited to finally present an early version of it. Would love feedback, especially from people who ship agent-based products. What’s the biggest part of your stack right now?