Skip to main content
Category: Research · Version: 1.0.0 · Author: ajaycodesitbetter · Compatibility: All targets (claude, codex, opencode, gemini, anti-gravity, openclaw, hermes)
LinkedIn Job Post to Buyer Pain Map — Signal Decoder for GTM Teams Turn LinkedIn job posts into an actionable buyer pain map with signal strength, urgency, and outreach angles for each account.

Install

What It Does

  • Parses 1–15 pasted job descriptions and groups them by company
  • Extracts team, seniority, responsibility emphasis, requirement language, and tool/stack hints from each post
  • Infers primary and secondary operational pains with cited evidence from the job text
  • Scores each account on 3 dimensions: signal strength (1–10), urgency (1–10), ICP fit (1–10)
  • Computes an overall priority score (10–100) using a transparent weighted formula
  • Determines buy-vs-build posture and company stage from hiring language
  • Generates 1–3 actionable outreach angles per account with talk-track bullets
  • Produces a structured handoff object ready for outreach-sequence-builder consumption

Requirements

Setup

Fill in GEMINI_API_KEY (required). No other dependencies. The skill runs entirely through agent instructions and Gemini API calls.

How to Use

Analyze a single job post:
Batch analysis of multiple posts:
Focus on positioning angles:
Handoff to outreach:

Input Format

Each job post should include at minimum:
  • Company name and job title (required)
  • Job description text — the full pasted content (required)
  • Location, seniority, team/function, URL (optional — inferred if missing)
Posts can be pasted as plain text, numbered lists, or structured JSON objects.

Output

For each company analyzed, the skill produces:
  1. Summary — headline, overall score, score breakdown with explanations, stage guess, buy-vs-build posture
  2. Buyer Pain Map — primary and secondary pains, each with label, description, and cited evidence
  3. Outreach Angles — 1–3 angles with narrative and talk-track bullets
  4. Handoff — structured block for outreach-sequence-builder (account summary, key pain, personas, tone)
Reports are saved to docs/pain-maps/ as dated markdown files.

Scoring Model

Overall Score = round((0.4 × signal + 0.3 × urgency + 0.3 × icp_fit) × 10) → 10–100 scale. See references/scoring-rubric.md for full scoring rules and modifiers.

Plays Well With

When NOT to Use

  • Need to find companies hiring → use linkedin-hiring-intent-scanner or yc-intent-radar-skill
  • Need to write outreach messages → use outreach-sequence-builder
  • Need to monitor platforms for signals → use reddit-icp-monitor or twitter-GTM-find-skill
  • Need contact data or email enrichment → this skill does not provide that

Project Structure

License

MIT