omkarpimpalgave208 / evidence-first-project-planner

Evidence-First Project Planner

When to reach for this

Use this skill when someone wants to evaluate a software or AI product idea, research competitors, plan an MVP, choose affordable tools, or validate an idea before investing time and money.

How to do it

1. Understand the idea

Identify the target user, the problem, the proposed solution, and the user's budget and timeline. Ask a clarifying question only when essential information is missing.

2. Research existing alternatives

When research tools are available, investigate existing products, open-source projects, and manual workarounds. Prefer reliable, current sources and provide source URLs.

Never invent competitors, prices, statistics, citations, or customer feedback. If research is unavailable, clearly say what remains unverified.

3. Separate evidence from assumptions

Label important findings as verified evidence, user-provided information, inference, or hypothesis. Explain which claims still need testing.

4. Find differentiation

Compare existing alternatives. Identify a specific unmet user need supported by evidence. If the idea has no clear advantage, suggest narrowing or changing it.

5. Design a small MVP

List only the essential features needed to test the core value proposition. Give each feature a measurable acceptance criterion. Separate essential features from future improvements.

6. Choose practical tools

Recommend a suitable technology stack based on the user's skills, budget, and hardware. Prefer free or open-source tools when appropriate. Explain known limitations and potential costs.

7. Plan three validation experiments

Where feasible, propose:

• A test to verify the user problem.
• A prototype usability test.
• A test measuring whether the solution achieves its intended outcome.

For each test, explain the hypothesis, procedure, metric, and success threshold. Mark proposed thresholds as hypotheses, not established facts.

8. Identify risks

Identify the most important technical, financial, privacy, security, and adoption risks. Provide a practical mitigation or experiment for each major risk.

9. Create an execution plan

Organize the work into milestones with deliverables, dependencies, and completion criteria.

Required output

1. Idea summary
2. Target user and problem
3. Evidence and unknowns
4. Existing alternatives
5. Differentiation hypothesis
6. MVP features and acceptance criteria
7. Suggested technology stack and costs
8. Three validation experiments
9. Risks and mitigations
10. Milestones
11. Next three actions
12. Decision: Build, Test Further, Pivot, or Stop

Watch out for

• Fabricated evidence, statistics, citations, or testimonials.
• Treating assumptions as verified facts.
• Adding too many features before validating the problem.
• Recommending tools that exceed the user's budget or hardware constraints.
• Claiming an idea is unique without researching alternatives.
• Giving vague success criteria.

If research tools are unavailable, disclose that limitation and do not pretend research was performed.

Example requests

• "Help me validate my startup idea before I build it."
• "Research competitors for my AI product."
• "Define the smallest useful MVP for my software idea."
• "How can I test whether customers need this product?"

Search keywords

software planning, startup validation, competitor research, MVP, product strategy, idea validation

Final quality check

Before responding, ensure that the target user and problem are specific, research claims are traceable, the MVP is small enough to test, success criteria are measurable, and uncertainties are clearly explained.

End with a clear decision and the next practical action.