discovery-process
by deanpetersdiscovery-process is a structured workflow for turning a vague product problem into a validated direction through framing, interviews, synthesis, and experiments. It helps product managers and UX researchers validate assumptions, clarify pain points, and decide what to build next.
This skill scores 78/100, which means it is a solid listing candidate for directory users who want a structured discovery workflow. The repository shows a real, reusable process for taking a product hypothesis through framing, interviews, synthesis, and experiments, so installing it should reduce guesswork compared with a generic prompt, though users should still expect to adapt it to their team and supporting skills.
- Clear triggerability: the frontmatter explicitly says it is for a full discovery cycle from problem hypothesis to validated solution, with concrete best-for scenarios like churn, onboarding, and continuous discovery.
- Good operational depth: the skill body is substantial, with 13 H2s, 29 H3s, and workflow/constraint signals that suggest a real step-by-step process rather than a placeholder.
- Useful install-decision value: the included template and example file show the intended output and a worked discovery path, helping agents understand how to execute it.
- No install command or companion support files are provided, so users may need to wire the workflow into their own skill stack manually.
- The repository appears focused on process orchestration rather than self-contained execution assets, so it likely works best when paired with separate framing, interview, and synthesis skills.
Overview of discovery-process skill
The discovery-process skill is a structured workflow for taking a vague product problem and turning it into a validated direction through framing, interviews, synthesis, and experiments. It is best for product managers, UX researchers, and cross-functional teams that need more than a generic brainstorming prompt and want a repeatable discovery-process for UX Research and product decisions.
What this skill is for
Use discovery-process when you need to test whether a problem is real, understand why it exists, and decide what to do next before building. It helps you move from “we think users struggle with X” to evidence-backed next steps.
Who gets the most value
This discovery-process skill fits teams working on retention, churn, activation, onboarding, or unclear customer pain. It is especially useful when stakeholders want confidence, but the team does not yet have enough evidence to justify a build.
What makes it different
The main value is sequence: the skill is designed to connect problem framing, interview planning, synthesis, and solution testing instead of treating them as separate one-off prompts. That makes it stronger than asking an LLM for “research ideas” or “user interview questions” in isolation.
How to Use discovery-process skill
Install and read the right files first
Run the discovery-process install step with the repository’s skill loader, then open skills/discovery-process/SKILL.md first. Next, inspect template.md and examples/sample.md to understand the expected output shape and how a complete cycle is documented.
Give it a real discovery brief
The skill works best when your input includes the problem area, audience, business context, and what decision depends on the result. A weak prompt says “help us with onboarding”; a stronger one says “activation dropped 15% for self-serve SMB users, and we need to know whether the main issue is comprehension, setup friction, or lack of value.”
A practical prompt pattern
Use the discovery-process usage pattern like this:
Run discovery-process for UX Research on our onboarding drop-off problem. Audience: first-time SMB admins. Goal: identify the biggest friction point, draft interview questions, propose a synthesis structure, and suggest one testable solution hypothesis.
If you already have evidence, include it. If you do not, say so explicitly. The skill is more useful when it knows whether it should start from a hypothesis, existing data, or a stakeholder question.
What to read and reuse
Start with the examples folder for a full cycle reference, then mirror the template for your own notes. The discovery-process guide is strongest when you reuse its phases as a working artifact: problem framing, research plan, interviews, synthesis, opportunities, experiments, and decision.
discovery-process skill FAQ
Is discovery-process better than a normal prompt?
Yes, when you need a repeatable process rather than a single answer. A normal prompt can generate interview questions or a summary, but discovery-process helps coordinate the sequence and keeps outputs tied to a decision.
When should I not use it?
Do not use it if the problem is already validated and you only need implementation details. It is also a poor fit for purely internal, non-user-facing questions where discovery interviews and synthesis would not add much value.
Is this useful for beginners?
Yes, if you can describe the problem clearly. Beginners get the most value when they follow the template and fill in the concrete context instead of asking the skill to invent the research question for them.
Does it fit UX Research workflows?
Yes. The discovery-process for UX Research angle is one of its strongest use cases, especially for interview planning, synthesis structure, and turning findings into an experiment or decision memo.
How to Improve discovery-process skill
Start with a sharper problem statement
The best results come from inputs that separate symptom, suspected cause, and business impact. For example: “Users abandon setup after connecting a bank account; we suspect they do not understand the next step, and this may be driving activation loss.” That is better than “users are confused.”
Add constraints that change the research plan
If you have a deadline, limited access to users, or a narrow segment, say so upfront. Constraints change whether the skill should recommend switch interviews, support-ticket review, lighter synthesis, or faster concept testing.
Ask for outputs you can execute
To get better discovery-process usage results, request concrete deliverables: an interview guide, a synthesis framework, opportunity statements, and a testable hypothesis. If you only ask for “insights,” the output is usually too abstract to move into a real decision.
Iterate after the first pass
Use the first output to tighten the next round. If the framing is too broad, narrow the audience. If the interview plan feels generic, add behavioral evidence or a specific funnel stage. If the proposed solution is premature, ask for more emphasis on problem validation before ideation.