Skip to content

template-variables.md

Latest commit

 

History

History
104 lines (80 loc) · 4.05 KB

template-variables.md

File metadata and controls

104 lines (80 loc) · 4.05 KB

Template Variables Reference

This document describes the variables available in the follow-up agent's system prompt template (linkedin/templates/prompts/follow_up_agent.j2) and the data structures from LinkedIn's Voyager API.

Agent System Prompt Variables

The follow-up agent template receives these named variables (not a spread profile dict):

Variable Type Description Example
self_name string The logged-in user's name (from /in/me/ marker) "Jane Doe"
product_docs string Product/service description from Campaign
campaign_objective string Campaign goal from Campaign
booking_link string Calendar link from Campaign "https://calendly.com/your-link"
full_name string Lead's first + last name "John Smith"
headline string or null Lead's profile headline "VP of Engineering at Acme"
current_company string or null Company from first position "Acme Corp"
location string or null Location as displayed on profile "San Francisco, California"
messages_exchanged integer Total number of messages exchanged with this lead (inbound + outbound) 5

Voyager API Profile Structure

The profile data parsed by linkedin/api/voyager.py contains the following fields. These are stored in Lead.profile_data (JSONField) and used internally for enrichment and qualification, though only a subset is passed to the agent template.

Top-Level Fields

Field Type Description
first_name string First name
last_name string Last name
full_name string First + last name combined
headline string or null Profile headline / tagline
summary string or null The "About" section text
public_identifier string or null LinkedIn handle (URL slug)
url string Full LinkedIn profile URL
location_name string or null Location as displayed on profile
geo dict or null Structured geographic info
industry dict or null Industry info
positions list of dicts Work experience entries
educations list of dicts Education entries
connection_degree int or null Connection degree (1 = connected, 2 = 2nd, 3 = 3rd)
connection_distance string or null Raw distance value from the API
urn string LinkedIn internal URN identifier

Positions

Each entry in the positions list:

Field Type Description
title string Job title
company_name string Company name
company_urn string or null LinkedIn URN for the company
location string or null Position-specific location
description string or null Role description text
date_range dict or null Start/end dates (see below)
urn string or null LinkedIn internal URN for this position

Educations

Each entry in the educations list:

Field Type Description
school_name string School or university name
degree_name string or null Degree type
field_of_study string or null Field/major
date_range dict or null Start/end dates (see below)
urn string or null LinkedIn internal URN

Date Range

Position and education entries may have a date_range dict:

{
  "start": {"year": 2020, "month": 3},
  "end": {"year": 2024, "month": 12}
}
  • start and end are dicts with year (int or null) and month (int or null).
  • A null end means the position is current.

Geo and Industry

The geo and industry fields are dicts with API-specific keys:

geo.defaultLocalizedNameWithoutCountryName  →  "San Francisco Bay Area"
industry.name                                →  "Computer Software"

Null Safety

Many fields can be null. In Jinja2 templates, use the default filter or conditional checks:

{{ headline | default("a professional") }}
{% if positions %}Current company: {{ positions[0].company_name }}{% endif %}