blog(ai): Run any coding agent in a sandbox, with one chat() call#1016
Conversation
Announce the TanStack AI sandbox layer + acpCompatible: coding-agent harnesses (Claude Code, Codex, OpenCode, Grok Build, or any ACP agent) run inside isolated sandboxes across local/Docker/Daytona/Vercel/Cloudflare and stream back through chat().
Deploying with
|
| Status | Name | Latest Commit | Preview URL | Updated (UTC) |
|---|---|---|---|---|
| ✅ Deployment successful! View logs |
tanstack-com | 20f4860 | Commit Preview URL Branch Preview URL |
Jun 30 2026, 01:54 PM |
📝 WalkthroughWalkthroughA new blog post ( ChangesNew blog post: coding agents in a sandbox
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~5 minutes Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Warning Tools execution failed with the following error: Failed to run tools: Ping-pong health check failed Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
@coderabbitai review |
✅ Action performedReview finished.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
🧹 Nitpick comments (1)
src/blog/run-coding-agents-in-a-sandbox.md (1)
116-120: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueConsider array-based command or escaping note for the ACP example.
The
commandtemplate literal builds a shell string:`pi --acp -m ${model} --cwd ${harnessCwd}`. IfharnessCwdcontains spaces, the resulting command splits incorrectly. Readers copying this pattern may hit a subtle bug.Consider showing an array form (if the API supports it) or adding a brief note about shell escaping:
// If the API accepts an array of arguments instead of a shell string: command: ({ model, harnessCwd }) => ['pi', '--acp', '-m', model, '--cwd', harnessCwd],Or add a inline comment warning about spaces in paths.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/blog/run-coding-agents-in-a-sandbox.md` around lines 116 - 120, The ACP example in acpCompatible uses a shell-interpolated command string, which can break when harnessCwd contains spaces. Update the command callback in the pi example to use an argument-array form if supported, or add a clear note/comment about shell escaping and quoting for model and harnessCwd so readers do not copy a fragile pattern.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Nitpick comments:
In `@src/blog/run-coding-agents-in-a-sandbox.md`:
- Around line 116-120: The ACP example in acpCompatible uses a
shell-interpolated command string, which can break when harnessCwd contains
spaces. Update the command callback in the pi example to use an argument-array
form if supported, or add a clear note/comment about shell escaping and quoting
for model and harnessCwd so readers do not copy a fragile pattern.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 66639043-2354-4568-bdde-b9f781123e35
⛔ Files ignored due to path filters (1)
public/blog-assets/run-coding-agents-in-a-sandbox/header.pngis excluded by!**/*.png
📒 Files selected for processing (1)
src/blog/run-coding-agents-in-a-sandbox.md
2 participants
Announces the TanStack AI sandbox layer + acpCompatible.
Adds a blog post (
src/blog/run-coding-agents-in-a-sandbox.md,library: ai) covering:chat()+withSandbox().acpCompatible— theopenaiCompatibleof coding agents: plug in any ACP agent (gemini --acp, Cursor, Goose, Pi, …) with no dedicated package.All code samples are grounded in the real API. Header image added at
public/blog-assets/run-coding-agents-in-a-sandbox/header.png.Summary by CodeRabbit