Agent Skills: CRM Log Activity

CRM Log Activity

Logging outreach activities (calls, messages, emails)

When to use

• After any outreach
• "log what I wrote"
• "record a call"
• For effectiveness analytics

Paths

| What | Path | |------|------| | Activities | $CRM_PATH/activities.csv | | People | $CRM_PATH/contacts/people.csv |

Schema

activity_id,linkedin_url,date,channel,activity_type,message_preview,result,response_quality,next_followup_date,audience_segment,hook_type,notes

How to log

import pandas as pd
from datetime import date, timedelta
import uuid

activities = pd.read_csv('$CRM_PATH/activities.csv')

new_activity = {
    'activity_id': f'act-{uuid.uuid4().hex[:8]}',
    'linkedin_url': 'https://linkedin.com/in/example',  # or email
    'date': str(date.today()),
    'channel': 'telegram',  # linkedin/email/twitter/telegram/whatsapp
    'activity_type': 'dm',  # dm/connect_request/followup/email/call/research
    'message_preview': 'Hi! This is Ivan from WeLabelData...',  # first 100 characters
    'result': 'sent',  # sent/replied/no_response/accepted/rejected
    'response_quality': '',  # positive/neutral/negative/meeting
    'next_followup_date': str(date.today() + timedelta(days=7)),
    'audience_segment': 'training_course',
    'hook_type': 'course_invitation',
    'notes': ''
}

activities = pd.concat([activities, pd.DataFrame([new_activity])], ignore_index=True)
activities.to_csv('$CRM_PATH/activities.csv', index=False)

After logging -- update the person

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

mask = people['linkedin_url'] == 'https://linkedin.com/in/example'
people.loc[mask, 'status'] = 'contacted'
people.loc[mask, 'last_contact_date'] = str(date.today())
people.loc[mask, 'next_followup_date'] = str(date.today() + timedelta(days=7))
people.loc[mask, 'last_updated'] = str(date.today())

people.to_csv('$CRM_PATH/contacts/people.csv', index=False)

Channels (channel)

• linkedin -- LinkedIn DM
• email -- Email
• telegram -- Telegram
• whatsapp -- WhatsApp
• twitter -- Twitter/X

Activity types (activity_type)

• dm -- Direct message
• connect_request -- Connection request
• followup -- Follow-up contact
• email -- Email message
• call -- Phone call
• research -- Research (not outreach)

Results (result)

• sent -- Sent, waiting for response
• replied -- Received a response
• no_response -- No response (after N days)
• accepted -- Accepted request/invitation
• rejected -- Declined

Response quality (response_quality)

• positive -- Interested
• neutral -- Neither yes nor no
• negative -- Not interested
• meeting -- Meeting scheduled

Batch logging

For mass outreach -- log multiple at once:

sent_to = ['email1@test.com', 'email2@test.com', 'email3@test.com']

for email in sent_to:
    new_activity = {
        'activity_id': f'act-{uuid.uuid4().hex[:8]}',
        'linkedin_url': email,  # using email as ID
        'date': str(date.today()),
        'channel': 'telegram',
        'activity_type': 'dm',
        'result': 'sent',
        # ...
    }
    activities = pd.concat([activities, pd.DataFrame([new_activity])], ignore_index=True)

Related skills

• telegram-send -- before logging
• update-lead -- update person status