JTBD Extractor

Elicitation Entry Rule

BEFORE asking anything in any module:

1. Read knowledge/elicitation-state.yaml
2. Find the corresponding module
3. If status == "covered" → summarize what exists, ask if user wants to go deeper
4. If status == "in_progress" → continue from current_probe
5. If status == "not_started" → start with story probe (see knowledge/story-probes.md)
6. After each question → update elicitation-state.yaml
7. After each story → generate SNAP-# (see Story Snapshots in ideate SKILL.md)

PURPOSE

Extract the real progress the actor is trying to make, not the feature they asked for.

This skill exists because users often describe a desired solution while the underlying job remains implicit. Its output should be written into the Opportunity Pack as JTBD-# entries linked to ACT-#.

OUTPUT CONTRACT

For each JTBD found, capture:

• type: functional | emotional | social
• actor: ACT-#
• trigger
• situation before
• anxiety or friction
• progress desired
• current alternative
• confidence: high | medium | low

EXTRACTION PROMPTS

Use these prompts one at a time, story-first:

• "Before thinking about the solution: what was this person trying to do at that moment? What would make them say 'it was worth it'?"
• "What changed that made this matter now? Tell me about the specific moment it became urgent."
• "Walk me through what feels risky or frustrating in how they do it today. What's the worst part?"
• "What do they use today instead? How did they end up with that workaround?"
• "If we nailed this, what progress would they actually feel — not just functionally, but emotionally?"

Refer to knowledge/story-probes.md for additional story-based probes per module. Refer to knowledge/vague-response-escalation.md for handling vague or generic answers.

ANTI-PATTERNS

• Do not write JTBDs as features
• Do not confuse persona description with the job itself
• Do not stop at the functional job when emotional or social jobs are relevant
• Do not accept "save time" without specifying where and how