Advanced Jobs-To-Be-Done (JTBD) framework for product analysis. A set of Claude Code skills that apply JTBD methodology to product decisions — segmentation, positioning, messaging, and validation.
Every recommendation is grounded in real user jobs, not assumptions, features, or demographics.
Foundation — product-jtbd skill with principles, anti-patterns, and 7 reference documents covering job story format, segment analysis, value propositions, landing page structure, friction analysis, evidence grounding, and product slop detection.
15 skills organized by workflow stage:
| Stage | Skill | What it does |
|---|---|---|
| Setup | teach |
Collects product context, writes Product Context section |
| Diagnostics | audit |
Full JTBD audit across 6 dimensions with severity ratings |
| Diagnostics | critique |
Strategic critique across 8 dimensions |
| Analysis | extract-jobs |
Extracts job stories from descriptions, feedback, reviews |
| Analysis | map-segments |
Groups users by shared jobs, scores segments |
| Analysis | find-friction |
Friction audit of user journey with severity scale |
| Analysis | score |
Scores and ranks segment opportunities |
| Refinement | sharpen |
Sharpens positioning and value prop with numbers |
| Refinement | distill |
Simplifies to ONE job, ONE segment, ONE value delta |
| Refinement | clarify |
Rewrites messaging through JTBD lens |
| Refinement | validate |
Finds riskiest assumptions, designs RAT tests |
| Generation | landing |
Generates landing page block structure |
| Generation | objections |
Creates evidence-based objection handlers |
| Generation | job-stories |
Writes full JTBD-format job stories |
claude install-skill https://github.com/bluzir/indispensable source/skills/product-jtbd
claude install-skill https://github.com/bluzir/indispensable source/skills/teach
claude install-skill https://github.com/bluzir/indispensable source/skills/audit
claude install-skill https://github.com/bluzir/indispensable source/skills/critique
claude install-skill https://github.com/bluzir/indispensable source/skills/extract-jobs
claude install-skill https://github.com/bluzir/indispensable source/skills/map-segments
claude install-skill https://github.com/bluzir/indispensable source/skills/find-friction
claude install-skill https://github.com/bluzir/indispensable source/skills/score
claude install-skill https://github.com/bluzir/indispensable source/skills/sharpen
claude install-skill https://github.com/bluzir/indispensable source/skills/distill
claude install-skill https://github.com/bluzir/indispensable source/skills/clarify
claude install-skill https://github.com/bluzir/indispensable source/skills/validate
claude install-skill https://github.com/bluzir/indispensable source/skills/landing
claude install-skill https://github.com/bluzir/indispensable source/skills/objections
claude install-skill https://github.com/bluzir/indispensable source/skills/job-storiesInstall product-jtbd first — it's the foundation that other skills reference.
- Install the skills (see above)
- Run
/teachto set up product context - Run
/auditfor a full diagnostic - Follow the recommended skills from the audit report
"I have a product, is the positioning right?"
teach -> audit -> sharpen -> clarify
"I have user feedback, what should I build?"
teach -> extract-jobs -> map-segments -> score -> find-friction
"I need a landing page"
teach -> extract-jobs -> map-segments -> sharpen -> objections -> landing
"Is my strategy sound?"
teach -> critique -> validate -> distill
- Job — Progress a person is trying to make in a specific context. Not a task, not a feature request.
- Segment — Group defined by shared jobs and contexts, not demographics.
- Value delta — Measurable Before/After difference your product delivers, benchmarked against market median.
- Product slop — Generic, ungrounded output that looks like product thinking but isn't. Detected via a 10-criterion scoring model.
- RAT — Riskiest Assumption Test. Lightweight experiment to validate the assumption that would be most costly if wrong.
MIT