GitHub PR Review Agent with OpenCode
A confident verdict on code nobody showed it is the mistake this template exists to avoid. An item about error handling can turn on a function three files away that the diff never touches, and the only honest answer is cannot tell with the reason attached. OpenCode leaves the engine open, so how reliably that honesty holds comes down to the model you connect, and swapping it takes a minute.
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 OpenCode
Preselected for this page — connect your OpenCode 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 OpenCode for this agent
Refusing beats inventing
Every wrong not met costs an author a rebuttal and costs your reviewers their trust in the whole comment. A model that admits the diff cannot settle an item is worth more here than a model that always answers.
Try a different model on Monday
The checklist, the credential and the wiring all stay put when you change the engine behind them, so putting two models against one repository costs a setting rather than a rebuild.
Before you fork
Anyone can read a comment on a public repository. What keeps internal detail out?
The instructions rule out internal ticket references, customer names and roadmap detail in anything posted on a public repository, and the single case where a fork edits release workflows or build scripts is routed to Slack rather than into the public thread. That is written policy rather than a hard boundary, so replay the staged fork contribution first and read the comment it produces with your own eyes.
Run it with a different agent
Put GitHub PR Review Agent to work on OpenCode
Free to start. Guided setup, a test run against staged conversations, and it's live.