Triage P2 bugs into draft PRs with demo videos
A Dot connected to GitHub and Slack watches #bugs, scopes and tests a fix in a cloud environment, and opens a draft PR with a short demo video.
Done withDots
- What they did
- A Dot, which is OpenAI's persistent agent with its own cloud Linux machine and browser, is connected to a GitHub repo and a Slack workspace. A custom rule tells it to investigate P2 bugs. It watches #bugs, spots recurring patterns, scopes a fix, writes and tests the patch in its cloud environment, and opens a draft PR with a short demo video. A person reviews and merges.
- How it went
- The source calls this the most compelling use case OpenAI demonstrated, but gives no success rates or benchmarks. The live DevDay Dot failed to respond on stage, and OpenAI warns Dots can make mistakes even when following rules.
- Worth knowing
- Audit token scopes before connecting: one developer's supposedly minimal GitHub token turned out to allow broader access, and the Dot used it. The first Dot needs Pro ($100/month and up) or Business Premium.
Try it yourself with Dots
Connect to GitHub repo [repo] and Slack channel [#bugs]. When a bug of priority [P2 or lower] is reported, scope the fix, test it in a cloud environment, and open a draft PR with a short demo video of the fix working. Post the PR link back in the Slack thread, and never merge anything yourself; stop once the draft PR is open and tell me what you verified.
1 more account of this
Source: byteiota.com · Oct 2, 2026
Five of these in your inbox every morning
The best things people got an AI agent to do, each with the prompt to try it.
More like this
Plan a trip and book flight, hotel and shuttle with an always-on agent
A user had Dots plan a trip and remind them to book the shuttle; the agent filled booking forms but kept asking for confirmation and wouldn't enter 2FA codes, so it only got partway.
Catch and prepare an unsent invoice for approval
An early tester's dot noticed an invoice to a publication was unsent and prepared it for approval.
Turn customer feedback into tested pull requests
In OpenAI's own launch example, a Dot watches feedback for recurring requests, scopes small fixes, builds and tests them, and returns PRs with demo videos.
- Claude
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Write and debug code across eight test tasks
A reviewer had Manus build a currency converter, fix a find_max bug and make a landing page, with mostly strong results.
Build a Pong game and website from scratch
A tester asked Devin to create a basic Pong-style game and a website, and it finished both in under 20 minutes.