agentleFS
peterbamuhigirepeterbamuhigire/chwezi-dev-engine/skills/product-business/product-discovery/SKILL.md

product-discovery

Skill31 starsChanged 19 days agoMITView on GitHub

Use when testing whether a product, feature, or workflow deserves investment through risk framing, prototypes, research, and build-pivot-kill decisions.

24 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use product-strategy.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • yannickYamoSkill

    product-strategy

    Build product strategy that drives real choices. Use when defining where to play and how to win, creating strategy canvases, defining ICP/Anti-ICP, positioning products, setting pricing strategy, planning GTM motions, or structuring strategic bets. Part of the Modern Product Operating Model collection.

    Also in 4 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. 0 are common, 13 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 22 edited versions we read keep these.

  • 13

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Default to read-only analysis and drafting.

    line 30
  2. Distinctive:

    If evidence, stakeholder decisions, specialist tooling, or authoritative commercial data are unavailable, deliver a labelled draft, …

    line 34
  3. Distinctive:

    Use the four risks to decide what needs evidence, then choose the cheapest prototype or test that can answer the risk.

    line 68
  4. Very specific:

    Use enterprise-ux-process when the work has enterprise constraints, stakeholder churn, or requirement ambiguity.

    line 90
  5. Very specific:

    Use the SRS engine's canonical decision-frontier-elicitation reference when open product decisions have prerequisites or several owners.

    line 91
  6. Distinctive:

    Before recruiting or building, run a lightweight market preflight: define the problem and target segment, inspect demand and alternatives, …

    line 95
  7. Distinctive:

    Every product must clear four risks before full delivery investment:

    line 125
  8. Distinctive:

    Before committing to deeper discovery, capture:

    line 138
  9. Distinctive:

    If these are unclear, the opportunity is not ready.

    line 145
  10. Distinctive:

    Ask about current behavior, not speculative future wishes.

    line 149
  11. Distinctive:

    Keep the target market narrow.

    line 150
  12. Distinctive:

    Use prototypes and manual concierge work before large build commitments.

    line 152
  13. Distinctive:

    Run a 1-2 week cycle:

    line 167
  14. Distinctive:

    Test with real users.

    line 172
  15. Distinctive:

    Use ARRIVE's Reframe stage as a hard checkpoint: the client signs off on the new frame before any concept work begins.

    line 193

“Common” is measured: kept by at least 6 in 10 of the 22 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Product Discovery

Acknowledgement: Shared by Peter Bamuhigire, techguypeter.com, +256 784 464178.

Required Inputs

InputRequiredUse
Decision, audience, and deliverableyesBound the business outcome
Source evidence, constraints, and owneryesGround recommendations and accountability
Approved budget, customer data, or production artefactsconditionalSupport high-impact execution

Skill-specific inputs:

  • Business objective, target market, and the product or feature idea under review.
  • Available research, usage data, constraints, and stakeholder context.

Capability and permission contract

Default to read-only analysis and drafting. Do not publish, send, price, promise, alter customer records, commit budget, or modify production artefacts without explicit authority and a named approver. Minimise confidential data, preserve provenance, and keep reversible copies.

Degraded mode

If evidence, stakeholder decisions, specialist tooling, or authoritative commercial data are unavailable, deliver a labelled draft, checklist, or decision memo. State what was not verified and do not claim approval, publication, financial accuracy, or customer acceptance.

Decision rules

ConditionActionStop condition
Output creates a commercial, customer, or delivery commitmentObtain named approval before releaseAuthority or terms are unclear
Evidence supports a reversible draftProduce it with assumptions and ownerRequired evidence conflicts
Tooling or data is incompleteSpecify validationA final executable artefact is expected

Domain Anti-Patterns

  • Inventing customer evidence, prices, benchmarks, or approvals. Fix: cite the source or mark the gap.
  • Publishing or sending a draft without authority. Fix: retain draft status and name the approver.
  • Hiding assumptions inside polished prose. Fix: expose them beside each affected decision.
  • Polishing presentation while the decision remains unclear. Fix: resolve audience, owner, and acceptance criteria.
  • Treating unavailable tooling as passed validation. Fix: record the unassessed check.
<!-- dual-compat-start -->

Use When

  • Evaluating a new product idea or market opportunity.
  • De-risking a major feature before it enters the backlog.
  • Deciding whether to pivot or persevere on an existing product direction.
  • Running a time-boxed discovery sprint with a cross-functional team.

Do Not Use When

  • The work is already defined at feature level and only needs implementation sequencing.
  • The request is a narrow screen-level UX refinement with no product-risk question.

Workflow

  • Use the four risks to decide what needs evidence, then choose the cheapest prototype or test that can answer the risk.
  • End with a concrete recommendation and the evidence supporting it.
  • For premium products, test willingness to pay, proof threshold, perceived quality, decision process, and buyer commitment before treating positive feedback as demand.

Quality Standards

  • Discovery must produce decisions, not theater.
  • Success metrics must be defined before deeper investment.
  • Prototype fidelity should stay as low as possible while still answering the risk question.

Anti-Patterns

  • Treating output volume as discovery progress.
  • Asking customers about hypothetical future preferences instead of present behavior.
  • Using colleagues or stakeholders as stand-ins for target users without recording the limitation.

Outputs

  • Opportunity assessment, discovery brief, test plan, prototype recommendation, evidence summary, and build / pivot / kill recommendation.

References

  • Use enterprise-ux-process when the work has enterprise constraints, stakeholder churn, or requirement ambiguity.
  • Use the SRS engine's canonical decision-frontier-elicitation reference when open product decisions have prerequisites or several owners.
  • Question-answering prototypes

Evidence ladder and scale gate

Before recruiting or building, run a lightweight market preflight: define the problem and target segment, inspect demand and alternatives, and distinguish search/interest signals from observed behaviour. Then escalate evidence only as the decision requires: interviews/observation, a concept or landing test, a click-through or workflow prototype, concierge/manual delivery, and finally real use, payment, renewal, or another material commitment. Predeclare the threshold, guardrail, time window, denominator, and persevere/pivot/kill rule.

For SaaS or product-scale decisions, do not treat acquisition or a polished prototype as product-market fit. Require a named activation event, repeat behaviour/retention evidence, a viable delivery and support path, and capacity to maintain the product/design system before recommending scale. If an AI-assisted workflow is used, preserve source context, record decisions, verify research against the original product/source, and review scoped prototype deltas before adoption.

Practitioner cross-checks: Eleken product-idea validation, SaaS launch, startup scaling, and AI design workflow. Use as prompts, not as proof of market size, outcomes, or tool performance.

Evidence Produced

CategoryArtifactFormatExample
CorrectnessProduct decision evidence boardMarkdown tableproblem, segment, evidence, assumption, option, experiment, threshold, and decision
<!-- dual-compat-end -->

Based on Cagan (2017) INSPIRED: How to Create Tech Products Customers Love, 2nd ed.

The 4 Product Risks

Every product must clear four risks before full delivery investment:

RiskQuestionValidated By
ValueWill customers choose to use or buy this?Demand tests, interviews
UsabilityCan customers figure out how to use it?Prototype testing
FeasibilityCan we build it with our team, tech, and time?Engineering spike
Business ViabilityDoes it work for the business?Stakeholder review

For premium offers, add a fifth commercial proof question: will the right buyer pay a premium now, and what proof, risk reduction, service layer, or authority asset is required for that decision?

Opportunity Assessment

Before committing to deeper discovery, capture:

  1. Business objective
  2. Key results
  3. Customer problem
  4. Target market

If these are unclear, the opportunity is not ready.

Customer Discovery Rules

  • Ask about current behavior, not speculative future wishes.
  • Keep the target market narrow.
  • Recruit real reference customers early when possible.
  • Use prototypes and manual concierge work before large build commitments.

Prototype Selection

Choose the cheapest prototype that answers the risk:

Prototype TypeBest For
Feasibility prototypeTechnical uncertainty
User prototypeUsability and task flow
Live-data prototypeValue and workflow realism
Hybrid prototypeMixed risk across value, usability, and feasibility

Discovery Sprint

Run a 1-2 week cycle:

  1. Frame the opportunity.
  2. Interview target users.
  3. Prototype 2-3 directions.
  4. Test with real users.
  5. Make a build / pivot / kill recommendation.

Companion Skills

  • enterprise-ux-process for feature-definition inside enterprise delivery environments.
  • feature-planning once the opportunity passes discovery.

ARRIVE: separating Research from Reframe

Many discovery processes collapse "research" and "synthesis" into one phase, which is exactly the failure mode that produces decks nobody acts on. The ARRIVE framework treats them as distinct stages — Research is ethnographic gathering, Reframe is the deliberate scope-and-meaning shift that names what to build against.

Load references/arrive-framework.md for:

  • The two-R front end (Research, then Reframe — never collapsed).
  • The Persona Needs Matrix and the Reframed Project Vision template that locks in scope.
  • The Concept Board (A) for low-fidelity concepts that test the idea, not the polish.
  • The Assumption Map (importance × evidence) that prioritises which assumptions to validate first.
  • Strategic Suitability as a fourth dimension alongside Desirability/Viability/Feasibility.

Use ARRIVE's Reframe stage as a hard checkpoint: the client signs off on the new frame before any concept work begins. Without that gate, ideation becomes unanchored.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.