Skip to main content
command CX Codex

analyze-call — Claude Code command: sales call analysis

analyze-call breaks down a call recording or transcript into metrics, key moments, risks and action items for sales teams.

Apache-2.0 ★ 450 updated 2026-04-03

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/commands/analyze-call.md
Open on GitHub

What analyze-call does

The command takes a call recording (link or ID), processes the transcript and calculates metrics: talk ratio, sentiment, filler words. It scores conversation stages (intro, discovery, value, objection handling, next steps) using a built-in rubric, detects signals — risks like missing stakeholders or weak impact statements and opportunities like champion quotes or deal expansion indicators. Outputs a brief report with scores, quotes and CRM tasks.

The --focus option sets analysis emphasis (exec alignment, technical depth, negotiation, storytelling), --compare benchmarks against a reference (top-performer, previous call, certification rubric). Flag --clips true generates a highlight reel reference list, --include-meddic true checks MEDDIC or BANT coverage. The command invokes sub-agents deal-analyst, conversation-engineer, call-strategist and skills call-analysis-framework, meddic-checklist; without them some functions will not work.

What it helps with

  • break down a call into metrics and key moments
  • assess MEDDIC or BANT coverage in a conversation
  • spot risks and opportunities from a call transcript
  • generate CRM tasks from call analysis
  • compare a call against a benchmark or previous one

Example prompts for the agent

  • Break down the call from s3://calls/demo-456 with focus on exec alignment and compare to top-performer
  • Analyze recording /analyze-call --recording "call-789" --include-meddic true --clips true
  • Score transcript transcript-123 against the rubric and give me CRM tasks

How to install analyze-call

Manually: put the files in place

.claude/commands/analyze-call.md

Call it with /analyze-call. For all projects use ~/.claude/commands/

Questions about analyze-call

How to install analyze-call in Claude Code?

Place the file in .claude/commands/analyze-call.md at the project root.

What does the analyze-call command do?

It processes a call transcript, calculates conversation metrics, scores stages against a rubric, identifies risks and opportunities, and outputs a brief report with quotes and recommendations for sales teams.

Which sub-agents does analyze-call need?

The command invokes deal-analyst, conversation-engineer, call-strategist and skills call-analysis-framework, meddic-checklist. Without them, some functions (signal interpretation, talk track critique, MEDDIC checks) will not work.

Source text

analyze-call.md · Apache-2.0
---
name: analyze-call
description: Breaks down a call recording/transcript into metrics, highlights, risks, and recommended actions.
usage: /sales-calls:analyze-call --recording "s3://calls/demo-456" --focus "exec alignment" --compare "top-performer"
---

# Command: analyze-call

## Inputs
- **recording** – required link or ID for the call.
- **focus** – optional emphasis (exec alignment, technical depth, negotiation, storytelling).
- **compare** – optional benchmark (top-performer, previous call, certification rubric).
- **clips** – true/false to generate highlight reels.
- **include-meddic** – true/false toggle to evaluate MEDDIC/BANT coverage.

## Workflow
1. **Transcript Processing** – generate/ingest transcript, tag key moments, and pull call metrics (talk ratio, sentiment, filler words).
2. **Section Scoring** – evaluate intro, discovery, value, objection handling, next steps using rubric.
3. **Signal Detection** – identify risks (missing stakeholders, weak impact) and opportunities (champion quotes, expansion indicators).
4. **Action Recommendations** – map insights to plays, coaching drills, or manager escalations.
5. **Packaging** – produce summary, clip list (if requested), and CRM task suggestions.

## Outputs
- Call analysis brief with scores, notable quotes, and next steps.
- Optional highlight reel references.
- MEDDIC/BANT coverage grid when enabled.

## Agent/Skill Invocations
- `deal-analyst` – interprets signals + coverage.
- `conversation-engineer` – critiques talk tracks and delivery.
- `call-strategist` – recommends storyline adjustments.
- `call-analysis-framework` skill – scoring rubric + output structure.
- `meddic-checklist` skill – ensures qualification completeness.

---

© gtmagents · Apache-2.0 · gtmagents/gtm-agents

Similar slash commands

All slash commands for Marketing and content →

From the same set224

The whole set →
launch-quantitative-survey The launch-quantitative-survey command for Claude Code plans quantitative surveys, supports fieldwork, and analyzes results through segmentation or scenario modeling. CX Codex +1 ★ 450repo stars launch-experiment launch-experiment turns an approved hypothesis into an instrumented experiment with guardrails and a rollout schedule. Use it when the design is signed off and the test needs to go live. CX Codex +1 ★ 450repo stars seo-writing seo-writing is a Claude Code skill that plans and drafts search-optimized articles connecting keyword intent with product positioning. It supports writer briefs and blog audits. CX Codex +1 ★ 450repo stars social-selling social-selling is a Claude Code skill for starting warm prospect conversations through LinkedIn and online communities. It guides social outreach that can turn interest in posts or webinars into meetings. CX Codex +1 ★ 450repo stars build-battlecard-suite build-battlecard-suite generates battlecards, objection-handling scripts, and enablement packs against competitors for sales or product teams. Formats output into decks, docs, Notion, or PDF. CX Codex +1 ★ 450repo stars update-api-reference The update-api-reference command for Claude Code compares API schemas with earlier versions and drafts docs, guides, and changelog updates as endpoints or parameters change. CX Codex +1 ★ 450repo stars run-win-loss-program run-win-loss-program runs a win/loss research cycle end to end, from picking deals to routing findings to product, sales and pricing. CX Codex +1 ★ 450repo stars run-market-landscape-study run-market-landscape-study — a Claude Code command that runs a full market research sprint: TAM/SAM/SOM sizing, trend aggregation, and competitor mapping. Outputs land as a deck, memo, or dashboard depending on the flag you pass. CX Codex +1 ★ 450repo stars

Project sponsors

Companies that keep this analytics open