the-theme-miner

Installation
SKILL.md

Boundary: This skill produces the research: themes, quotes, personas. For writing copy from that research, use the-campaign-composer, the-cold-opener, or the-page-shipper. For building audience segments, use the-lifecycle-mapper. For extracting the brand's own voice (not the customer's), use the-voice-fingerprint.

Context

  1. Check for .agents/product-context.md. If missing, ask the user to run /gtm:product-context first. If the user prefers to proceed without it, ask inline for: product one-liner and target ICP.
  2. Read references/customer-research-methods.md for the JTBD extraction framework, confidence scoring, sample bias checklist, and persona template.

Inputs

  1. Ask: "What do you already have?": transcripts, survey results, support tickets, reviews, win/loss notes, NPS responses, or nothing yet.
  2. If the user has assets: ask them to paste the raw text. Do not proceed on a summary of the material. Work from the actual quotes.
  3. If the user has nothing yet: ask for their ICP type (B2B SaaS, SMB, developer, B2C, enterprise) so research can be sourced from the right public channels using WebSearch/WebFetch: company review sites, public forums, and community discussions relevant to that ICP. State plainly that this is secondary research on public sentiment, not first-party customer data, and should be weighted as such.
  4. Ask: "What's the goal?": improve messaging, build personas, find product gaps, understand churn, or general synthesis. Ask: "What do you want delivered?": synthesis report, persona document, VOC quote bank, or competitive intel summary.

Process

Installs
4
GitHub Stars
9
First Seen
Jul 26, 2026