Business Bench

Tasks · build · sales

B2B sales CRM with accounts, contacts, a staged deal pipeline, weighted forecast, and activity log

crm-sales-pipeline · 3600 s per turn · source

The ask

We're a nine-person consulting firm and I want one place for our client accounts, their contacts, and our deals, with the calls, emails, and meetings reps log on a deal showing on that deal and on the client's account. Our accounts, contacts, and pipeline are attached; they were kept in three separate spreadsheets, so there are duplicates, and I want each account and each person in there only once. Deals go Lead, Qualified, Proposal, Negotiation, then Won or Lost: nothing goes to Proposal without an amount and a contact, nobody marks a deal Lost without a reason, and the dashboard should show our pipeline weighted at 10, 25, 50, and 75 percent for those four open stages. Each rep sees only their own accounts and deals, our analysts can read everything but change nothing, and I look after the users and settings. Give me an admin login and a login for Leila Haddad, one of our reps.

Seed files the agent must import: accounts.csv</code>, <code>contacts.csv</code>, <code>deals.csv. The agent must leave a RESULT.json with a reachable URL, an admin login, and the named restricted login.

Acceptance checklist

26 binary items. The first 14 are the enterprise baseline shared by every build task (delivery, invitations, roles, server-side authorization, scoped live dashboard, exact lists and exports, audit metadata, validation, persistence); the rest are application-specific. Core items, which together define "first usable": 1, 3, 5, 6, 8, 14, 15, 22, 25.

TagItems
Exact10
Permission3
Rule3
Dashboard2
Delivery2
Sharing2
API1
Audit1
Automation1
Persistence1

The full checklist with per-item "how to check" instructions and expected figures is in checklist.md. Testers see it; the agent never does.

Change requests

Each change is a new one-shot turn on the carried-forward workspace. Only the owner's text below is sent; the tester items are withheld. After each change the full checklist is rerun, so regressions count.

Change 1 · 3 new checklist items

I want a win-rate report: for each rep, how many deals they've won and lost and their win rate, plus the company total. Put the company win rate on the dashboard as well.

Change 2 · 3 new checklist items

When a deal is marked Won, create a kickoff task for the rep who owns it, due three working days after the close date. One kickoff task per deal: if a deal gets reopened and won again, don't make another.

Change 3 · 3 new checklist items

Add a Verbal commit stage between Negotiation and Won, weighted at 90%. Reps shouldn't be able to mark anything over $50,000 as Won themselves; it should come to me to approve.

No published score. Build runs need a human acceptance pass on a live URL, and no arm has completed one for the full set. Automated probes generate a tester sheet; they do not substitute for it. Results will be published as a separately labeled release when an arm completes.