Grading is where this agent can go wrong quietly, because the instruction is to grade only against your criteria and never against a name, a city or the profile sitting behind the form. With OpenCode the model is one field you control, so when a run starts scoring on impressions the fix is a model change rather than a rewrite of your qualification rules.
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.
Free solar quote - August
Full name: Amira Haddad, prefilled. Email: amira.haddad@gmail.con, typed. Phone: +971 50 447 2210, typed. Monthly electricity bill: AED 1,800 to 2,400. Do you own the property: yes, villa in Al Barsha. When are you looking to install: next month, before the tariff change. Submitted 3 minutes ago.
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect Meta
One sign-in. The agent acts through your account, scoped to what this template uses.
Connect HubSpot
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.
Qualification criteria
What makes a Meta lead worth calling and what rules one out, written against the questions your form actually asks, including which answers you refuse to take at face value when they arrive prefilled.
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.
A lead qualifies because an answer met a line you wrote. Nothing about who they appear to be enters the grade, and that is easy to audit after the fact.
When a prefilled answer contradicts a typed one, the note says which is which rather than picking a winner. Preserving the conflict is more useful to a caller than resolving it.
It does, one run per lead, so a campaign that suddenly performs costs more on the day it works. That is usually the trade you want, and the lever is the model rather than the volume: something small handles the repair and dedupe work fine, and you can hold a larger one for forms where the grading is genuinely close.
Free to start. Guided setup, a test run against staged conversations, and it's live.