Discovery5-point gate · Any stack
Map the riskiest assumption
Find the belief most likely to waste the loop, then test it before polishing anything else.
Use this when
A long risk list where everything is marked high priority
Technical uncertainty tested before uncertain customer demand
An experiment whose result cannot change the plan
The quality gate
- The desired customer and business outcome is stated before assumptions are listed
- Assumptions cover desirability, viability, feasibility, and usability
- Each assumption is scored separately for consequence and evidence strength
- Exactly one current riskiest assumption is selected with a written rationale
- A cheap falsification test has a threshold, deadline, and resulting decision
Prove it: Run the proposed test as a paper walkthrough: every possible result must map to pass, fail, or gather-specific-evidence-next, with no result interpreted as automatic success.
Install or copy
Save the download as riskiest-assumption-map/SKILL.md in your agent's supported skills directory, or copy the complete package below. Load it only when its description matches the work.
Portable SKILL.md
--- name: riskiest-assumption-map description: "Maps and prioritizes uncertain product assumptions, then designs the cheapest falsification test. Use before committing meaningful build effort or when a plan depends on several unproven beliefs." --- # Map the riskiest assumption Apply this skill during **Discovery** work. It supports Any stack. Find the belief most likely to waste the loop, then test it before polishing anything else. ## Instructions Map the assumptions behind this product decision. Begin with the customer outcome and business outcome. List assumptions under desirability, viability, feasibility, and usability. For each, score consequence if wrong and current evidence strength; do not use confidence as a substitute for evidence. Select exactly one riskiest assumption: the high-consequence belief with the weakest evidence that can still change the plan. Design the cheapest ethical test that could falsify it, with a deadline, numeric or observable threshold, and an explicit action for pass, fail, or ambiguous evidence. Defer lower-risk work until this test is read. ## Detect the problem - A long risk list where everything is marked high priority - Technical uncertainty tested before uncertain customer demand - An experiment whose result cannot change the plan ## Hold the gate Do not declare this phase complete until every item passes: - [ ] The desired customer and business outcome is stated before assumptions are listed - [ ] Assumptions cover desirability, viability, feasibility, and usability - [ ] Each assumption is scored separately for consequence and evidence strength - [ ] Exactly one current riskiest assumption is selected with a written rationale - [ ] A cheap falsification test has a threshold, deadline, and resulting decision ## Verify the result Run the proposed test as a paper walkthrough: every possible result must map to pass, fail, or gather-specific-evidence-next, with no result interpreted as automatic success.
More discovery skills
Want this quality gate applied to a real product decision?
Start a First Loop