Skip to content
GTM Agent Weekly

Generate weekly RevOps reports from live CRM data on a schedule

This agent pulls live CRM data on a schedule and sends ready-made pipeline, forecast and rep reports to chat or email. Revenue operations stops rebuilding the same spreadsheets every Monday.

Paste this into ChatGPT, Claude or your agent tool. See the prompt

  • RevOps Reporting
  • intermediate
  • RevOps
  • Sales leader
Category
RevOps Reporting
Difficulty
Intermediate
Est. cost
Not stated by source

Works with

Problem

Someone on RevOps spends Monday mornings exporting CRM data into spreadsheets and slides that go stale quickly.

Setup

Claude Code connects to the CRM over MCP, pulls live data on a schedule, computes pipeline health, forecast accuracy, rep scorecards, funnel conversion and enrichment metrics, and delivers formatted reports to Slack or email.

Prompt

Starter prompt by The GTM Agent Weekly, written from the setup described on this page. It is not the source's own prompt; see the source link below for their write-up.

Starter prompt · 23 lines
# ROLE
You produce our weekly RevOps report from live CRM data and deliver it formatted.

# INPUTS
- {{CRM}}: opportunities, activities, leads, owners (read access).
- {{DEFINITIONS}}: how we define each metric: pipeline created, coverage, win rate, cycle
  time, forecast accuracy, funnel stages.
- {{PERIOD}}: the reporting window. {{DELIVERY}}: Slack channel or email list.

# TASK
1. Pipeline health: open pipeline by stage and owner, created this period, coverage vs. target.
2. Forecast accuracy: last period's forecast vs. actual, by team.
3. Rep scorecards: activities, meetings, pipeline created, wins.
4. Funnel conversion between stages for the period.
5. Data quality: share of records missing required fields.
6. Three notable changes vs. the previous period, in plain sentences.

# OUTPUT
A formatted report with tables and a short summary at the top.

# GUARDRAILS
- Use {{DEFINITIONS}} exactly. Show the query or filter behind each number in an appendix.
- Read-only. If data looks wrong (e.g. negative values, big jumps), flag it instead of reporting it as fact.

Replace every {{PLACEHOLDER}} before running. Works with any capable LLM and runner.

Tools

Steps

  1. Connect the CRM to Claude Code over MCP.
  2. Define the reports to automate and the metric definitions behind each.
  3. Schedule runs that pull live data and compute the metrics.
  4. Deliver formatted reports to Slack or email.

Results

Guide from an enrichment vendor; no outcome data reported.

What goes wrong

Gotchas

  • Reports built on CRM fields full of blanks mislead; the source pairs this with its own enrichment product, a vendor interest to note. (editor note)
  • Metric definitions must be written down explicitly or the agent infers them from field labels. (editor note)
  • Validate automated reports against a manually built one before trusting them unattended. (editor note)

Source and credit

syncgtm.com ›

Kushal Magar, SyncGTM · vendor

Source dated 2026-04-27 · last reviewed 2026-10-11. Summarized in our own words; see the source for details.

Related.

Test whether an ROI walkthrough should be a required deal stage

This analysis checks whether deals that include an ROI or business-case conversation actually win more often, comparing like with like. Leaders can decide on a required sales step with evidence instead of instinct.

Works with any of SalesforceHubSpotGong +6

  • RevOps Reporting
  • advanced
  • Reference setup