User Stories & Job Stories (grounded in customer evidence)

Turn a feature ($ARGUMENTS) into a set of independent, testable stories where each story cites the customer quote that motivates it, and every acceptance criterion reflects a stated customer expectation — not an invented one. This is the difference between a backlog of guesses and a backlog you can defend.

Step 0 — Relevance & availability

Confirm this is customer-facing product work and Evermuse is present. If the MCP isn't connected, produce the stories from the templates but label them ⚠ ungrounded and tell the user to authorize the MCP. (See ${CLAUDE_PLUGIN_ROOT}/skills/using-evermuse/SKILL.md, Step 0.)

Evermuse Grounding (required)

Follow ${CLAUDE_PLUGIN_ROOT}/skills/using-evermuse/SKILL.md.

Required — one parallel batch of searches. Verify the product (Rule 1), then fire in a single parallel batch:

• 3–4 evidence searches, each worded from a different angle (limit up to 50). Evidence comes back rich and varied — expect large, useful result sets.
• one guidance search and one context search (limit up to 50). These are usually sparse or empty; run them anyway and note when they're thin.

Read each response's digest — it reports how many more results exist. Use judgment on whether a query is worth pulling deeper (raise limit toward the 100 max and/or page with next_offset to avoid repeats), weighing payload size, remaining context, task complexity, and the value of the data. For deep pulls, consider spawning sub-agents — instruct them to return every citation with the same metadata the tools return (url, who_said_it, meeting_name, created_at) so you can still cite.

Optional — considered use. Once grounded, reach for the other tools only when they add value: a quote-angled search for verbatim voice, read_source for a single deep dive, inline citations (references/citations.md), and add_source to save the deliverable. Sources, citations, and saving are optional — not required.

Instructions

1. Pick the format that fits. Both are valid; choose per the user's ask (offer to switch):
   • User story — role-centric: "As a [role], I want [action], so that [benefit]." Follows the 3 C's (Card / Conversation / Confirmation).
   • Job story — situation-centric (JTBD): "When [situation], I want to [motivation], so I can [outcome]." Reach for this when the trigger/context matters more than the role, which the evidence often reveals.
2. Derive stories from the evidence, not the design alone. Group quotes and needs by distinct user situations/jobs; each cluster becomes one story. Replace the source skill's reliance on "$DESIGN + $ASSUMPTIONS" with evidence as the primary driver — designs and assumptions refine stories, but the corpus decides which stories exist and how they rank.
3. Respect INVEST. Independent, Negotiable, Valuable, Estimable, Small, Testable. Split any story too big for one sprint.
4. Write acceptance criteria that mirror stated expectations. Each AC should trace to something a customer expressed — a needed behavior, an edge case they hit, a performance bar they complained about. 4–6 criteria per story; observable and testable.
5. Rank by demand. Order stories by how strongly the evidence backs them (mentions across accounts). Flag any story that's a guess (no evidence) as assumption — validate.
6. Chain onward. Stories feed test-scenarios (Given/When/Then from these ACs) and write-feature-spec. Reuse this grounding rather than re-searching (credits — Rule 7).

Story templates

User story

Title: [feature name] As a [role], I want [action], so that [benefit]. 1 Design: [link] Acceptance Criteria: 1–6 observable, testable criteria — each tracing to a stated expectation.

Job story

Title: [job outcome] When [situation], I want to [motivation], so I can [outcome]. 1 Design: [link] Acceptance Criteria: 6–8 outcome-focused criteria (situation recognized, motivation enabled, feedback visible, edge cases handled).

Each story block ends with its inline citation.

Further reading

• Adapted from phuryn/pm-skills (MIT) — merges user-stories + job-stories.
• How to Write User Stories · JTBD Masterclass