GitHub PR Review Agent with OpenClaw
Almost everything in this brief is a prohibition. No approving, no merging, no closing, no pushing, no editing a label or a reviewer, and where a human has already reviewed, every point that review made has to be dropped and the comment has to open by naming whose work it builds on. OpenClaw is a simple agent aimed at a simple job, and the credential cannot do the forbidden things regardless.
Watch it work before it's live
Run a staged conversation — no account needed. The agent handles it for real while a simulated world answers its tool calls; nothing touches real accounts, and nothing is actually sent.
Stream tile responses instead of buffering the whole body
Openopened by @marijke-devos
Replaces the buffered writer in the tile handler with a streaming encoder, so a 40MB vector tile stops sitting in memory before the first byte goes out. Nine files, 214 lines added, 118 removed, all under internal/tiles. Adds a benchmark showing peak memory falling from 96MB to 11MB on the Jakarta extract, plus table tests for the partial write path. The description links issue 402 and states the wire format is unchanged.
Set up in minutes
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect GitHub
One sign-in. The agent acts through your account, scoped to what this template uses.
Connect Slack
One sign-in. The agent acts through your account, scoped to what this template uses.
Review checklist
What a pull request has to satisfy in this repository, put the way you would explain it to somebody reviewing here for the first time: which items are hard blockers, which paths you treat as sensitive, and what you want asked when the diff alone cannot settle one.
Runs on OpenClaw
Preselected for this page — connect your OpenClaw account during setup, or switch to NoClick's built-in models with one click.
Watch it handle a test run
A staged conversation against a simulated world — then it’s live.
Why OpenClaw for this agent
Second reader, not a repeat
Going second means saying so and then adding what is left. An agent that restates the three points a human already made is the fastest way to have its comments filtered out.
The credential is the real fence
Repository read plus permission to write issue comments is the whole grant. GitHub does not count an issue comment as a review, so required approvals and branch protection behave exactly as you configured them.
Before you fork
A teammate reviews while the agent is still writing. What happens?
It reads existing reviews at the start of the run, so a review landing halfway through will not be in what it saw and the comment can repeat a point somebody just made. In practice it fires within a minute of the pull request opening, usually well ahead of a person, and the worst case is a duplicated sentence rather than a wrong verdict.
Run it with a different agent
Put GitHub PR Review Agent to work on OpenClaw
Free to start. Guided setup, a test run against staged conversations, and it's live.