# Buyer Job Intent Analysis

Recover source-bound buyer jobs, struggling moments, desired progress, forces...

- Repository: https://github.com/elvisun/newsjack (1668 stars)
- Category: Design
- License: MIT
- Last updated: 2026-10-25
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add elvisun/newsjack --skill buyer-job-intent-analysis -g -a claude-code -y`
Codex: `npx skills add elvisun/newsjack --skill buyer-job-intent-analysis -g -a codex -y`
Cursor: `npx skills add elvisun/newsjack --skill buyer-job-intent-analysis -g -a cursor -y`
Gemini CLI: `npx skills add elvisun/newsjack --skill buyer-job-intent-analysis -g -a gemini-cli -y`
Claude Code plugin marketplace: `/plugin marketplace add elvisun/newsjack`

## Description

Recover source-bound buyer jobs, struggling moments, desired progress, forces, workarounds, information acts, journey states, criteria, constraints, roles, locales, and authentic language. Use on approved ICP hypotheses plus customer, search, review, forum, procurement, support, or public-market evidence before designing an AI-visibility prompt architecture.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Buyer Job Intent Analysis

Recover what people are trying to accomplish and how they express it. Do not turn product features into imagined demand.

This skill inherits the ethical floor from `skills/ETHICS.md`. It enforces anti-hallucination, permission, provenance, and decay-aware handling of current market evidence. Anti-spray and human-send are not applicable.

## Inputs

Require:

- approved `icp_hypotheses.json`;
- `source_manifest.json` with permission and provenance;
- any user-supplied transcripts, queries, reviews, support material, or market sources.

Research missing public-market language when permitted. Prefer evidence in this order:

1. lawfully collected relevant AI conversations with collection metadata;
2. customer or prospect interviews and calls;
3. on-site search, support, chat, sales, and win/loss evidence;
4. paid-search, Search Console, marketplace, and site-search queries;
5. public reviews, forums, communities, RFPs, procurement guides, and competitor reviews;
6. broad search and People Also Ask proxies;
7. company copy;
8. LLM expansion.

Public conversational corpora may inform style, turn count, and multilingual naturalness. They must not supply category prevalence or copied prompts.

## Grade evidence

| Grade | Meaning | Eligible use |
| --- | --- | --- |
| `A` | Direct relevant behavior or verbatim customer/prospect language with provenance | Wording and exposure-weight inputs |
| `B` | Credible public-market behavior or search proxy with provenance | Wording with an explicit proxy label |
| `C` | Company assertion or expert hypothesis | Research hypothesis; approval required |
| `D` | LLM-generated expansion without independent support | Rotating discovery only |

A count is a count within the supplied corpus. Never relabel it market frequency.

## Extract jobs and language

For each supported ICP, extract:

The rest: https://github.com/elvisun/newsjack/blob/HEAD/plugins/newsjack/skills/buyer-job-intent-analysis/SKILL.md