Pull request incoming!
AI for coding
Hyper writes features, runs tests, and opens pull requests, with the context of your whole team: the issue, the discussion, the docs, and what happens in production.
What Hyper does with code
- Writes and edits code. Hyper edits projects, writes features, and runs tests in its own cloud environment, where it can also use a browser.
- Takes on big jobs. For a task with many steps, Hyper divides the work among parallel agents and keeps going, summarizing earlier context to make room.
- Works in your repositories. In GitHub, Hyper opens, reviews, approves, and merges pull requests, creates branches, and commits files. It also works with merge requests in GitLab.
- Reads the code that it needs. Hyper can read the code, commits, branches, and Actions runs in the repositories where you install the Hyper GitHub app.
Code with the context of your team
Hyper sees more than your code. It also knows why the code changes, because it searches the tools where your team works:
- The issue and its comments in Linear or Jira
- The discussion in Slack and in pull request reviews
- The spec and the docs in Notion, Confluence, or Google Drive
Debug with production data
Hyper can check what really happened before it changes a line. It searches logs, traces, and metrics in Datadog, Grafana, and Logfire, reads traces in LangSmith, and queries your databases in Neon and Supabase.
What to ask Hyper
- “Fix the bug in the issue that Sam filed today, write a test for it, and open a pull request.”
- “Why did checkout errors go up last night? Check Datadog and the pull requests merged yesterday.”
- “Review the open pull request for the billing change, and leave comments on anything risky.”
- “How many people signed up this week? Run it on our Neon database.”
You stay in control
By default Hyper asks before it commits, merges, or changes anything in your tools, and it always asks before it commits files from your computer. You can approve a kind of action for good, or ask for a draft first.