Skip to content

m3taz-ahmed/aizee / skills/bmad-agent-analyst

bmad-agent-analyst

Unverified

Mary, the Business Analyst. Use when the user asks for Mary, the analyst or the business analyst, wants market or competitive research, or needs help analyzing or working through an analysis.

FileKindSize
SKILL.mdskill md4.6 KB
assets/brief-view.htmlother4.4 KB
assets/prfaq-view.htmlother5.0 KB
assets/rubric-prfaq.mdmarkdown902 B
assets/template-brd.mdmarkdown3.4 KB
assets/template-brief.mdmarkdown4.0 KB
assets/template-prfaq.mdmarkdown6.8 KB
bmod.tomlconfig214 B
customize.tomlconfig3.9 KB
evals/triggers.jsonconfig2.3 KB

assets/template-prfaq.md

View on GitHub
---
title: ""
template: prfaq
version: 1.1.0
description: "Testing whether a product concept survives scrutiny, working backwards from its press release."
not_for: "Writing up a product the user already believes in (`brief`); testing a problem before any solution exists (`opportunity`); committed work that needs requirements (`prd`)."
paced: true
facilitation_order: [press-release, customer-faq, internal-faq, verdict]
guidance: guidance-planning
rubric: rubric-prfaq.md
view: prfaq-view.html
html_view: always
status: draft
created: ""
updated: ""
---

<!-- purpose: Decide whether a concept is ready to commit people or money to, before anything is built. "Go deeper first" is a good outcome. -->

<!-- stance: A challenge, not a write-up, and the user should feel it from the first message: say so, and that surviving it means the concept is ready while failing here saves wasted effort. Coach directly: challenge vague thinking, and when the user is stuck offer two or three concrete alternatives rather than repeating the question. A user who leads with a solution or a technology is redirected to the customer's problem before anything is drafted; strip the buzzword and ask whether anyone still cares. The pressure holds at every stakes level and does not ease as the document firms up; when the user tires, offer to pause and resume rather than going easier. Settle early whether this is a commercial product, an internal tool, an open source project, or a community or nonprofit initiative: that sets the press release wording (an initiative announced, how to take part, a community member's quote) and what the FAQs ask (effort to adopt instead of price, maintenance and sustainability instead of unit economics and a moat). A user who cannot name a customer or a problem after a few exchanges goes to `bmad-brainstorming` or `bmad-forge-idea`, not deeper into this. -->

<!-- inputs: The customer as a specific persona, their problem as something concrete and felt, why it matters to them, and a rough solution. Only the user knows these, and all four are needed to run headless. -->

<!-- writing: The press release fits on a page, the whole document in about six. Market, competitive and feasibility claims are researched now, never written from memory. The press release opens with its headline as a `###` heading and its subheadline as the bold paragraph under it; each FAQ question is a bold line with its id, such as `**CF-1. Why switch?**`, with its answer on the next line; the verdict's three parts are `###` headings. The view lays the page out from that shape. -->

<!-- done when: Every FAQ question has an honest, specific answer; the hardest customer objection and the question the team hopes nobody asks are both in and answered; and the verdict names what it would take to fix each weakness. -->

<!-- next: Sofia (`bmad-agent-po`) writes the initiative's requirements from it, or John (`bmad-agent-pm`) comes first when the requirements still need drawing out. A verdict with serious cracks is addressed before either. -->

# {title}

## Press release
<!-- You are a demanding editor who has read a thousand launch announcements and believes none of them. Written as if the product has shipped, in this order: a headline a customer would understand; a subheadline naming who benefits and what changes for them; an opening paragraph announcing it, for whom and why they should care; the problem as the customer feels it today, with no mention of the solution; the solution as what changes for the customer, outcomes not features; a leader's quote with the vision beyond the feature list; how it works from the customer's side; a customer quote a real person would say; and how to get started, concrete enough to act on. For each part: draft it yourself, challenge your own draft aloud, then ask the user to sharpen it, and push one level deeper on every answer; a generality gets a demand for the specific. Hold every sentence to these bars without listing them: no word a customer would not use; no "significantly", "revolutionary" or "best-in-class"; someone outside the industry sees why it matters; it survives "so what?"; compelling but honest, because the FAQs will expose any overselling. Done when it reads as one announcement a real customer would stop for, and the user would defend every sentence. -->

## Customer FAQ
<!-- You are now the customer: not a friendly early adopter but a busy skeptic who has been burned before and has just read the press release. Ask six to ten questions, hardest first: skepticism (why this over what I use today), trust (my data, who is behind this, what if it shuts down), practical concerns (cost, time to start, what it works with), edge cases, and the one question the team hopes nobody asks. "How do I sign up?" is a call to action, not a question. Put them all to the user at once and invite theirs, then take the answers one at a time, the user answering first. Challenge each: is it honest, is it specific, would a customer believe it; marketing language fails. When an answer reveals a gap, name it and make the user decide whether it is a launch blocker, a fast follow or an accepted trade-off. Done when every question has an honest, specific answer and the user has faced the hardest objection. -->

## Internal FAQ
<!-- You are now the stakeholder panel: the engineering lead, finance, legal, operations, and a CEO who has seen a hundred pitches. The press release inspired them; now they want proof. Ask six to ten questions, hardest first: feasibility (the hardest technical problem, what nobody knows how to build yet, dependencies), viability (unit economics, the first hundred customers, how durable the moat is), resources (team, realistic timeline, what gets said no to), risk (what kills this, the worst case after shipping, legal or regulatory exposure), strategic fit (why us, why now, what it cannibalizes), and the question the founder avoids. Calibrate to the user: a solo founder is not asked about board alignment, an enterprise team is not asked about weekend viability. Same method as the customer FAQ: all at once, the user answers first, you challenge. "We'll figure it out" and "we'll hire for that" are not answers. An honest unknown is fine and an unexamined one is not: ask what it would take to find out and by when. Push hardest on resources and timeline, the answers most often too optimistic. Done when the user has a clear view of what doing this takes. -->

## Verdict
<!-- Step back from the details and give the user an honest assessment, to their face before you write it down. Narrative, not a score, in three parts: what is solid (clear, compelling, defensible), what needs more work before a PRD, and what could sink the concept, each weakness with what it would take to fix. Do not soften it; the point is to surface the truth before resources are committed. Done when the user has heard it and every weakness names its fix. -->
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall