What	Path
Companies	`$CRM_PATH/contacts/companies.csv`
People	`$CRM_PATH/contacts/people.csv`
Clients	`$CRM_PATH/relationships/clients.csv`
Partners	`$CRM_PATH/relationships/partners.csv`
Leads	`$CRM_PATH/relationships/leads.csv`
Products	`$CRM_PATH/products.csv`

Schemas

contacts/companies.csv

company_id,name,website,linkedin_url,type,industry,geo,size,description,created_date,last_updated

type: company, enterprise, ngo, individual
size: small, medium, enterprise, individual

contacts/people.csv

person_id,first_name,last_name,email,phone,linkedin_url,company_id,role,notes,created_date,last_updated

relationships/clients.csv

client_id,company_id,product_id,status,contract_start,contract_end,mrr,currency,primary_contact_id,notes,created_date,last_updated

status: active, paused, churned

relationships/partners.csv

partner_id,company_id,product_id,partnership_type,status,since,primary_contact_id,revenue_share,notes,created_date,last_updated

partnership_type: training_partner, workforce_partner, reseller_agreement, referral_partner

relationships/leads.csv

lead_id,company_id,product_id,stage,source,source_direction,source_detail,priority,primary_contact_id,estimated_value,currency,next_action,next_action_date,notes,created_date,last_updated,last_contact_via_primary

stage: new, qualified, proposal, negotiation, won, lost
source: google, website, facebook, linkedin, email, telegram, referral, direct, event, calendly, research
source_direction: inbound, outbound
source_detail: (optional) campaign name, referrer, post URL, etc.

How to add

1. New company

import pandas as pd
from datetime import date

df = pd.read_csv('$CRM_PATH/contacts/companies.csv')

# Check for duplicate
if 'example.com' in df['website'].values:
    print("Company already exists!")
else:
    new_row = {
        'company_id': 'comp-example',
        'name': 'Example Inc',
        'website': 'example.com',
        'type': 'company',
        'industry': 'Technology',
        'geo': 'USA',
        'created_date': str(date.today()),
        'last_updated': str(date.today())
    }
    df = pd.concat([df, pd.DataFrame([new_row])], ignore_index=True)
    df.to_csv('$CRM_PATH/contacts/companies.csv', index=False)

2. New contact

# First make sure the company exists!
companies = pd.read_csv('$CRM_PATH/contacts/companies.csv')
if 'comp-example' not in companies['company_id'].values:
    print("Add the company first!")

people = pd.read_csv('$CRM_PATH/contacts/people.csv')

new_person = {
    'person_id': 'p-example-001',
    'first_name': 'John',
    'last_name': 'Doe',
    'email': 'john@example.com',
    'company_id': 'comp-example',
    'role': 'CEO',
    'created_date': str(date.today()),
    'last_updated': str(date.today())
}

3. New client (relationship)

# Company and product must exist!
clients = pd.read_csv('$CRM_PATH/relationships/clients.csv')

new_client = {
    'client_id': 'cli-example-001',
    'company_id': 'comp-example',
    'product_id': 'prod-labeling',
    'status': 'active',
    'contract_start': str(date.today()),
    'primary_contact_id': 'p-example-001',
    'created_date': str(date.today()),
    'last_updated': str(date.today())
}

Validation (REQUIRED!)

Before saving:

ID is unique
company_id exists in companies.csv (for people, relationships)
product_id exists in products.csv (for relationships)
person_id exists in people.csv (for primary_contact_id)
No duplicates (check by email, website, name)
created_date and last_updated are filled in

After changes:

Run change-review skill before PR

ID formats

Type	Format	Example
Company	comp-{name}	comp-acme
Person	p-{company}-{number}	p-acme-001
Client	cli-{company}-{number}	cli-acme-001
Partner	ptnr-{company}-{number}	ptnr-acme-001
Lead	lead-{company}-{number}	lead-newco-001

Related skills

update-lead -- update existing record
query-leads -- search
change-review -- review before PR

add-lead

CRM Add

When to use

Architecture

Paths