This skill is a smart router, not one giant method.
Use it to size the decision, pick the right lane, apply enough structure to reduce regret, and end with a clear commitment, review trigger, and next action.
Quick Access
| Need | Reference |
|---|---|
| Clarify values before comparing options | references/01-framing-and-options.md |
| Pick the right lane fast | references/00-lane-selection.md |
| Frame the decision and build options | references/01-framing-and-options.md |
| Run the right evaluation method | references/02-evaluation-lanes.md |
| Challenge assumptions and stress-test | references/03-stress-testing.md |
| Handle stakeholder disagreement | references/04-group-decisions.md |
| Turn analysis into commitment | references/05-commitment-and-review.md |
| Grab a framework card quickly | references/06-method-cards.md |
| Produce the final artifact | templates/decision-record.md |
Use This Skill For
- choosing between named alternatives
- deciding whether to commit now, wait, pilot, or stop
- breaking analysis paralysis
- structuring high-stakes or uncertain decisions
- documenting why a decision was made and what could change it
- group decisions where trade-offs or ownership are muddy
Do Not Use This As The Main Tool When
- the choice is trivial, low-stakes, and reversible
- the user only wants information, not a decision
- there is only one viable option and the real work is execution
- the situation is an immediate emergency with no time for deliberation
Operating Principles
- Match rigor to stakes, reversibility, and uncertainty.
- Include the baseline: do nothing, delay, or keep the current path.
- Prefer the lightest method that still protects against obvious error.
- Do not invent fake precision when evidence is thin.
- Separate facts, estimates, values, and politics.
- End with commitment, not endless optionality.
Default Workflow
1. Clarify values and what matters
Read references/01-framing-and-options.md first.
Before comparing options, name:
- what the user is actually optimizing for
- which values or principles should guide the choice
- which trade-offs are acceptable versus unacceptable
- which non-negotiables must survive any analysis
If these stay implicit, later scoring will look rigorous but rest on the wrong foundation.
2. Frame the decision
Read references/01-framing-and-options.md.
Turn the problem into a one-sentence decision with:
- a clear verb
- a decision owner
- a deadline or horizon
- a baseline option
3. Select the lane
Read references/00-lane-selection.md.
Classify the decision by:
- stakes if wrong
- reversibility
- time pressure
- uncertainty
- stakeholder complexity
Then choose the lane that fits.
4. Run the lane-specific method
Read references/02-evaluation-lanes.md.
If the user wants a compact menu of named frameworks, also read references/06-method-cards.md.
Common lanes:
- quick
- trade-off
- sequence
- value
- uncertainty
- dynamic
5. Stress-test before committing
Read references/03-stress-testing.md.
Apply at least one challenge pass for modest decisions and two or more for high-stakes decisions.
6. Resolve ownership or disagreement when needed
If the issue is really stakeholder conflict, unclear authority, or value mismatch, read references/04-group-decisions.md before forcing more analysis.
7. Commit and document
Read references/05-commitment-and-review.md and write the result in templates/decision-record.md shape.
Response Contract
Unless the user asks for another format, produce:
- the decision statement
- the lane chosen and why
- the values, non-negotiables, or principles guiding the choice
- the options considered, including the baseline
- the criteria or evaluation logic used
- the main uncertainties and stress-test results
- the chosen action, first steps, and review triggers
Routing Rules
- If the decision is reversible and low stakes, use the quick lane and move.
- If there are multiple alternatives across conflicting criteria, use the trade-off lane.
- If the value of waiting or piloting matters, use the sequence lane.
- If outcomes can be estimated credibly in costs and benefits, use the value lane.
- If future states dominate the choice, use the uncertainty lane.
- If the environment is moving fast and action creates feedback, use the dynamic lane.
- If people are disagreeing about values or ownership, use the group-decision reference first.
- If the user's values are unclear, do not rush into comparison; clarify what matters first.
Guardrails
- Do not overengineer two-way-door decisions.
- Do not under-analyze one-way-door decisions.
- Do not confuse fear of commitment with lack of information.
- Do not use a matrix to justify a choice already made.
- Do not reopen settled decisions without new evidence or a review trigger.