Everything this agent may touch is a list you wrote: tags that already exist, one assignee, one comment. Closing, deleting, archiving and rewriting the title all sit outside it, and a repeat request gets named rather than merged. OpenCode is a good match for a loop defined that tightly, because you choose the model and the instructions do the constraining.
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.
Landing page for the October 8 partner webinar
TO DOAcreated by Aisyah Rahman
We need this live by September 24 for the partner webinar on October 8. Copy and speaker headshots sit in the shared drive folder linked below, registration should post to HubSpot form 4412. Budget code is MKT-Q4-08.
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect ClickUp
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.
Intake rules
How your intake actually works - which list new requests land in, the tags you sort them into, what a request must contain before work can start, and which requests deserve a Slack ping.
Team map
Each type of request paired with the ClickUp member who should own it, plus the person who catches anything matching nothing on the list.
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.
It sorts into tags your space already carries instead of coining new ones, so six months of running does not leave you with forty near identical tags to clean up.
A page of intake rules asks more of the model than four lines does. OpenCode lets you move that dial without rebuilding the workflow around it.
No. After the first pass the task is no longer new and already carries the agent's own tag and comment, so both halves of the gate fail and the second event ends without a write. The case it cannot solve is two separate tasks describing the same request, which is why it names the sibling task in its comment and leaves both open for a person to merge.
Free to start. Guided setup, a test run against staged conversations, and it's live.