An AI-powered sales research assistant for gathering and organizing company and product intelligence.
Sales Brain is a Cursor-based workflow that helps you build a comprehensive sales intelligence package for your company. It uses AI to research, organize, and structure everything your sales team needs to sell effectively.
- Company profile - Your company's overview, mission, differentiators, target market
- Product documentation - Features, value props, competitive positioning for each product
- Buyer personas - Target roles, their pain points, goals, discovery questions
- Competitive battlecards - Head-to-head comparisons with competitors, when you win/lose
- Objection handling - Common objections with proven responses
- Sales plays - Actionable playbooks for specific selling situations
- Case studies - Customer success stories with metrics
- Founders building go-to-market materials from scratch
- Sales leaders creating enablement content for their team
- Sales enablement standardizing messaging and playbooks
- RevOps documenting sales processes and competitive intel
- Type
/startin Cursor - Provide your company name and website
- AI scrapes your public information and generates structured sales intelligence
- Review, refine, and add your proprietary insights
- Your sales team uses the output for prospecting, pitching, and closing
- Run the setup script:
./setup.sh- Activate the virtual environment:
source venv/bin/activate- Open the project in Cursor
In Cursor:
/start
Then follow the 11-phase workflow:
- Company research
- Product detection
- Target companies
- Personas
- Pain points
- Value propositions
- Use cases
- Competitive intelligence
- Objection handling
- Case studies
- Sales plays
- Final step: Auto-generate
INDEX.md(file inventory & loading rules) andREADME.md(folder overview) for the company
All generated content follows the templates in templates/ (see Templates below).
All generated content follows templates in templates/. Use them for consistency when adding or editing objects.
| Template | Used for |
|---|---|
company-template.md |
Company overview, mission, differentiators, scraped info |
target-companies-template.md |
ICP, company profiles by type, qualification |
product-template.md |
Product overview, features, value prop, competition |
persona-template.md |
Role, responsibilities, pain points, goals, buying process |
pain-points-template.md |
Pain points (description, impact, solution, key message), priority matrix, discovery questions, trigger events |
value-proposition-template.md |
Summary, problem/solution, value drivers, quantified value, ROI story, messaging framework, competitive positioning |
use-case-template.md |
Context, challenge (before/after), solution, value delivered, proof points, sales conversation guide |
competitor-template.md |
Overview, products, strengths/weaknesses, feature comparison, battlecard, objection handling |
objection-template.md |
Objections with acknowledge, respond, proof point, follow-up question; prevention; quick reference |
case-study-template.md |
Snapshot, executive summary, customer/challenge/solution/results, quotes, sales use, related objects |
sales-play-template.md |
Play overview, trigger events, qualification, execution steps, objection handling, resources |
Each company gets its own directory under brains/ with all sales intelligence files. After the full workflow, INDEX.md and README.md are auto-generated for that company.
sales-brain-for-cursor/
├── templates/ # Templates for all objects (shared)
├── brains/ # All company research lives here
│ ├── {company-slug}/ # Company-specific directory
│ │ ├── INDEX.md # Auto-generated: file inventory & loading rules for AI
│ │ ├── README.md # Auto-generated: human-friendly folder overview (GitHub default)
│ │ ├── company.md # Company overview
│ │ ├── scraped/ # Scraped JSON (main.json, products, case-studies, etc.); use before re-scraping
│ │ ├── products/
│ │ ├── target-companies.md
│ │ ├── personas/
│ │ ├── pain-points/
│ │ ├── value-propositions/
│ │ ├── use-cases/
│ │ ├── competitors/
│ │ ├── objections/
│ │ ├── case-studies/
│ │ ├── sales-plays/
│ │ └── scraping.log # URLs scraped (written when using -d)
│ └── {another-company}/
└── ...
Full Sales Brain examples (all 11 phases + INDEX/README) are in brains/apollo/ and brains/revenue-io/. Other examples: brains/salesloft/, brains/gong/, brains/lautie/.
Example: Revenue.io (recent full run):
brains/revenue-io/
├── INDEX.md # File inventory & loading rules for AI
├── README.md # Folder overview, proof points, quick links
├── company.md
├── target-companies.md
├── scraped/ # main.json, products.json, case-study-*.json, competitor-*.json, etc.
├── products/ # 11 products (revenue-platform, salesforce-dialer, moments, ...)
├── personas/ # 8 personas (sales-leader, account-executive, revops, ...)
├── pain-points/ # 8 persona pain-point files
├── value-propositions/ # 6 product–persona value props
├── use-cases/ # 5 use cases
├── competitors/ # Gong, Outreach, Salesloft
├── objections/ # common + sales-leader, revenue-operations, account-executive
├── case-studies/ # HPE, Nutanix, OrthoFX, FreshBooks, NFI Industries
└── sales-plays/ # Consolidation, Real-time coaching, Scale meetings
Note: Most commands require {company} parameter (e.g., salesloft).
| Command | Description |
|---|---|
/start |
Begin the full 11-phase workflow |
/continue {company} |
Continue workflow (e.g., /continue salesloft) |
/status |
List all companies with summary |
/status {company} |
Detailed status with freshness indicators |
/search {company} {query} |
Search all files (e.g., /search salesloft ROI) |
/refresh {company} |
Full refresh - re-scrapes all sources |
/generate-index {company} |
Generate INDEX.md (file inventory) |
/generate-readme {company} |
Generate README.md (folder overview) |
/generate-visualization {company} |
Generate interactive HTML graph |
/list all |
List all companies |
/list all {company} |
List all objects for a company |
/list products {company} |
List products |
/list personas {company} |
List personas |
/add product {company} |
Add a new product |
/add persona {company} |
Add a new persona |
/add pain-points {company} |
Add pain points for a persona |
/add value-prop {company} |
Add value proposition |
/add use-case {company} |
Add a use case scenario |
/add competitor {company} |
Add competitive intelligence |
/add objections {company} |
Add objection handling |
/add case-study {company} |
Add a customer success story |
/add sales-play {company} |
Add a sales playbook |
/update company {company} |
Re-research company info |
/update target-companies {company} |
Update target companies |
Generated files follow the templates in templates/. Summary by type:
- README.md — Auto-generated; company tagline, proof points table, differentiators, folder structure, quick links. Shown by default on GitHub.
- INDEX.md — Auto-generated; file inventory by category, loading rules for AI agents, topic → file mapping. Use for context loading.
Use README.md for a one-page overview; use INDEX.md for full file list and when to load which file.
Company overview, industry, founded, headquarters, size, mission/vision, target market, key differentiators, website, scraped information (headings, links), additional notes.
Overview of target segments; company profiles (industry, size, characteristics, why they buy, common challenges); referenced customers; additional notes.
Company, overview, main features, problem solved, value proposition, target customers, competition table, pricing model, additional notes.
Role/title, responsibilities, company profile, key problems & pain points, goals & objectives, how our products solve their problems, buying process, communication preferences, additional notes.
Overview; pain points (description, impact, current solutions, frequency, severity, product solution, key message); pain point priority matrix; discovery questions; trigger events; quick reference table.
Summary, problem, solution; key value drivers (benefit, proof point, customer quote); quantified value table; ROI story; messaging framework (elevator pitch, email subject lines, key phrases, avoid); competitive positioning.
Overview; context (target company profile, primary persona, trigger event); challenge (current/desired state); solution (products, steps); value delivered (quantified + qualitative); proof points (customer examples, data); sales conversation guide (discovery questions, demo focus, talking points); related objects.
Overview, company info, product overview, positioning & messaging, strengths/weaknesses, feature comparison, competitive battlecard (when we win/lose, differentiators, their claims vs reality), handling competitive situations, trap-setting questions, pricing comparison, resources.
Objections with: why they say this, acknowledge, respond, proof point, follow-up question; objection prevention (questions to ask early, red flags); quick reference table.
Snapshot table; executive summary; the customer; challenge (situation, pain points, impact, what they tried); solution (why they chose us, products, timeline); results (metrics table, headline results, qualitative benefits); customer quotes; sales use (best for, talking points, objections this addresses); related objects; resources.
Play overview table; when to run (trigger events, ideal signals, qualification criteria); objective & value proposition; execution steps (research, outreach, discovery, demo, proposal); objection handling table; resources & tools; success metrics; related objects.
.cursor/rules/— Cursor AI rules (workflow, sales-brain scraping, templates).cursor/commands/— Slash commands (e.g.start.md,continue.md,add-*.md,generate-index.md)templates/— Markdown templates for every object type (company, product, persona, pain-points, value-proposition, use-case, competitor, objection, case-study, sales-play)
- Be specific when entering the company name
- Use the main website URL (not subpages)
- Review carefully before confirming - the AI will research based on your approval
- Add products manually if automatic detection misses some
- Check scraping.log to see which URLs were scraped
- Load README.md or INDEX.md when using AI agents for company context
Sales Brain uses a Python script for all web scraping (.cursor/rules/sales-brain/scripts/scrape.py). The workflow uses -d brains/{company-slug}/ so every scrape is logged to that company’s scraping.log.
# Scrape a URL (log to company directory)
python .cursor/rules/sales-brain/scripts/scrape.py scrape <url> -d brains/{company-slug}/
# Scrape homepage + 1 level of subpages (saves to scraped/ for later use)
python .cursor/rules/sales-brain/scripts/scrape.py scrape https://company.com -d brains/{company-slug}/ --subpages
# Scrape and save to JSON (e.g. for use in phases)
python .cursor/rules/sales-brain/scripts/scrape.py scrape <url> -d brains/{company-slug}/ -o brains/{company-slug}/scraped/main.json
# Load existing scraped data (check before re-scraping)
python .cursor/rules/sales-brain/scripts/scrape.py load-scraped -d brains/{company-slug}/
python .cursor/rules/sales-brain/scripts/scrape.py load-scraped -d brains/{company-slug}/ -p productsOptions:
-d, --log-dir— Directory forscraping.log(and where to find/save scraped data)-o, --output— Save JSON output to file (e.g.brains/{company-slug}/scraped/products.json)--subpages— Scrape homepage + one level of subpages; saves to companyscraped/for use in phases-f, --follow— Follow discovered links (about, products, pricing, contact, all)-m, --max-pages— Maximum pages when following links (default: 10)
Extracted data: Page title, meta description, headings (h1–h3), important links (about, products, pricing, contact), main text content.
Strategy: Use existing files in brains/{company-slug}/scraped/ when present (e.g. main.json, product or case-study JSON). Only run new scrapes for URLs not already covered or when scraped/ is empty. Scrape first, validate with user second.
The example company data in this repository (e.g. brains/apollo/, brains/revenue-io/, brains/salesloft/, brains/gong/, brains/lautie/) is for demonstration purposes only.
- All information was gathered from publicly available sources (company websites, public documentation, case study pages)
- This data illustrates how the tool works and what output to expect
- Example playbooks, battlecards, objection handling, and case studies are samples to demonstrate the framework
- This is not official sales material from any company mentioned
- For real sales use, generate your own research and add proprietary insights
The framework, templates, and tooling are the primary value of this project.
Tomas Zeman
📧 tomas.zeman@gmail.com
MIT