Create a PRD (grounded in customer evidence)

Produce a comprehensive but readable Product Requirements Document — the business case for $ARGUMENTS — where the problem, the objective, the segments, and the value propositions are each backed by something a real customer actually said, with a source link. A generic PRD reads plausibly; this one reads true, because the "why" comes from the corpus.

Pick the right sibling. create-prd (this skill) = a lighter 8-section business PRD for aligning stakeholders and leadership on the what/why. write-feature-spec = the rigorous, spec-driven sibling with prioritized user stories, Given/When/Then acceptance scenarios, and measurable success criteria for a team to build from. When unsure which the user wants, offer both.

Step 0 — Relevance & availability

Confirm this is customer-facing product work and the Evermuse tools are present. If the MCP isn't connected, produce the PRD from the template but label it ⚠ ungrounded — Evermuse not connected 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. Anchor each section in evidence, not assertion. Walk the 8 sections in references/prd-template.md. For each, ask what the corpus says before you write:
   • Background / Why now — what changed for customers? Lead with the pain, quoted. A quote-angled search on the problem.
   • Objective + Key Results — the objective maps to an evidenced customer outcome; the KRs measure reduction of a stated pain. (For a full OKR set, chain to brainstorm-okrs.)
   • Market Segment(s) — segments are defined by the problem/job customers described, not demographics. Name the accounts/personas the evidence actually came from.
   • Value Proposition(s) — each pain avoided / gain created is a real quote, not a guess. This is where the customer's voice does the most work.
   • Solution (UX, key features, tech, assumptions) — flag assumptions explicitly so the team can validate them.
   • Release — relative timeframes (now / next / later), never hard dates.
2. Reuse grounding across the session. If a spec, OKR set, or research brief already pulled this evidence, carry the quotes and source badges forward instead of re-searching (credits — Rule 7).
3. Write for a wide audience. Short sentences, minimal jargon; a PRD is read by leadership and engineers alike.
4. Working-Backwards variant (optional). If the user wants an Amazon-style press release / FAQ instead of (or before) the full PRD, use references/wwas-press-release-template.md. Its customer-quote sections must use real customer quotes from a quote-angled search, attributed — never invented testimonials.

Deliverable

An 8-section PRD (or the Working-Backwards press release) as clean markdown, with a Customer Evidence callout near the top summarizing themes + demand strength and inline citation badges throughout (links live inline — no Sources footer). Offer to save it (guidance).

Further reading

• Adapted from phuryn/pm-skills (MIT).
• PRD template · AI PRD template
• Rigorous sibling: ${CLAUDE_PLUGIN_ROOT}/skills/write-feature-spec/SKILL.md