Cold Email Reply Agent with OpenCode
One misclassification costs more than a hundred correct ones save: replying to an angry unsubscribe, or arguing with someone asking where you got their address. Classification quality is a model property, which makes OpenCode's model field the most important setting on this agent. The action side is narrow enough to stay safe either way.
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.
Re: cutting the returns backlog at Nordveld
This landed at a decent time, we are looking at exactly this problem for our Rotterdam warehouse. We handle around 400 returns a day and most of the work is still manual. What does it cost for one site, and could you do a call Thursday or Friday afternoon? I would want our operations manager on it as well.
Set up in minutes
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect Instantly
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.
Offer context
What you are selling and to whom - the pitch behind the campaign, the pricing you are happy to put in writing, what the product does and does not do, and the proof points a reply may use. The agent answers from this and nothing else.
Reply policy
How you want replies handled - the one next step you offer an interested lead, which questions the agent may answer alone, how an unsubscribe should be marked in Instantly, and the tone you want it written in.
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
Classification is the risk
Everything downstream follows from one word, so that is where model strength earns its cost.
Any language, any wording
A request to stop contacting us can arrive in any language and any phrasing, including one covering colleagues. Recognising it is not a keyword check.
Before you fork
How do I test a model on unsubscribes?
Use Angry Unsubscribe, where the reply is hostile, mentions a bought list, and asks that colleagues be left alone too. The only correct output is a lead update and a Slack line, with no reply sent at all. A model that composes an apology there has failed the test that matters most for your sending domains.
Run it with a different agent
Put Cold Email Reply Agent to work on OpenCode
Free to start. Guided setup, a test run against staged conversations, and it's live.