Morning deal-risk brief from yesterday's Gong calls and Attio deals
Every morning this agent reads yesterday's recorded calls, checks each deal in the CRM and posts a ranked list of deals most likely to slip to Slack. Managers see objections and competitor mentions before standup.
Paste this into ChatGPT, Claude or your agent tool. See the prompt
- Deal Review / Forecasting
- advanced
- Sales leader
- RevOps
- Category
- Deal Review / Forecasting
- Difficulty
- Advanced
- Est. cost
- Not stated by source
Problem
The warning signs on a slipping deal are usually in call recordings and the CRM, but nobody connects the two every day. Managers find out in the forecast call, when it is too late.
Setup
A scheduled agent runs each morning. It pulls the previous day's calls from Gong, analyzes each transcript for sentiment, objections and competitor mentions, and matches each call to its deal in Attio for stage, value and close date. A weighted formula combines sentiment, days to close, engagement and objections into a 0 to 1 risk score, deals are ranked, and the top ones are posted to Slack as a brief with stage, close date, the specific objections and the next steps discussed on the call.
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.
# ROLE
Each morning you review yesterday's sales calls and post a short deal-risk brief so managers
know which deals need attention today.
# INPUTS
- {{CALLS}}: yesterday's call transcripts with attendees (e.g. Gong).
- {{CRM}}: deals with stage, value, close date and owner (e.g. Attio).
- {{CHANNEL}}: where the brief goes (e.g. Slack). {{TOP_N}}: how many deals to show.
# TASK
1. Match each call to its deal by attendee emails. Skip calls with no deal and list them at the end.
2. For each call, note: buyer sentiment (positive, neutral, negative) with a quote, objections
raised, competitors mentioned, next steps agreed.
3. Rate each deal's risk high, medium or low using: negative sentiment, unresolved objections,
a close date coming up soon, and no agreed next step.
4. Rank by risk, then by deal value.
# OUTPUT FORMAT (Slack)
Deal risk brief, {{DATE}}
1. {{DEAL}} ({{OWNER}}) | {{STAGE}} | closes {{CLOSE_DATE}} | risk: high
Objections: ... | Competitors: ... | Next step: ... | Call: {{LINK}}
Calls with no deal: ...
# GUARDRAILS
- Quote evidence for sentiment and objections. No quote, no claim.
- Don't post call content in public channels for deals marked sensitive. Replace every {{PLACEHOLDER}} before running. Works with any capable LLM and runner.
Tools
Steps
- Connect the call recorder, CRM and Slack with credentials the agent can use on a schedule.
- Each morning, fetch the previous day's calls and their transcripts.
- Have the model pull sentiment, objections, competitor mentions and next steps from each transcript.
- Match each call to its CRM deal and read stage, value and close date.
- Score risk with a fixed weighted formula, rank the deals and post the top ones to a Slack channel before standup.
Results
Tutorial with no outcome data.
What goes wrong
Gotchas
- Gong takes 10 to 15 minutes to process a call, so runs too soon after a call miss its transcript.
- Calls that cannot be matched to a CRM deal drop out of the brief silently unless you report them.
- Agree the risk weights with your managers first, or the ranking will not match their judgment. (editor note)
Get 5 like this every week, free.
Real agent setups with sources and failure modes, in one short weekly email.
Free. One email a week. Unsubscribe anytime.
Source and credit
Scalekit · vendor
Source dated 2026-05-18 · last reviewed 2026-10-11. Summarized in our own words; see the source for details.