Repair & Reuse · Question 06
Lessons retrieval
Which previous decision and outcome is most relevant to the problem we face now?
Match the failure mechanism before borrowing the action. For the supplier-delay scenario, L01 is the relevant precedent: reserving a backup did not establish a one-week recovery; onboarding itself took two weeks. Pre-qualify the backup while the main order proceeds.
No current incident is identified in the supplied case. Supplier delay is the starting example, not a diagnosis. Choose any of the five supplied scenarios below.
Fictional what-if workspace. Changes stay in this tab and reset on reload; no record is saved and no external system is changed.
1 · Identify the current mechanism
Retrieve a previous decision
Try a short problem description
Optional deterministic lookup. It matches the mechanism words listed under “Retrieval rules”; it does not understand the situation. A mixed description returns candidates for you to choose.
2 · Read the outcome, not just the action
Context
Previous decision / action
Expected outcome
Actual recorded outcome
What this changes for the two-hub launch
3 · Test a proposed use
Does the mechanism carry across?
The lesson is a candidate for planning. Similar words alone do not make the old action suitable for this launch. The following checks are your hypothetical assumptions, not verified evidence.
No application simulated yet.
In-tab simulation result · no approval or saved record
- Retrieved evidence
- Mechanism judgment
- Assumed checks
- Outcome limit
- Next evidence
To correct this result, change the scenario or checks and simulate again. Reset clears the whole workspace. Reload restores the supplier-delay example. No purchase, staffing change, booking change or approval has occurred.
Five scenarios, five traceable cases
These are the complete supplied records. Use a row to inspect its outcome and test its application above.
| Case | Context | Action | Expected | Actual recorded | Lesson |
|---|
Retrieval rules and limits
The scenario selector uses an exact one-to-one mapping: supplier delay → L01; volunteer absence → L02; quality rework → L03; stakeholder approval → L04; demand surge → L05. There is one historical case per mechanism, so no ranking, confidence percentage or “best of many” claim is justified.
Description lookup dictionary
Case-insensitive whole-word/phrase matches after Unicode normalization. At least one term retrieves a candidate; extra matches do not increase authority. One candidate is selected automatically. Two or more require a choice; none leaves the current scenario unchanged. This is a deliberately small vocabulary, not semantic search.
Application rule: Different mechanism → do not transfer. Uncertain mechanism → investigate before applying. Same mechanism + any missing check → hold application. Same mechanism + both assumed checks → candidate for planning, still unapproved.
The historical outcome fields are observations, not proof that the action caused them. There are no dates, sample sizes, comparison groups, staffing rosters, contract terms, missed-appointment counts or onboarding details. Do not infer these. The scenario-specific transfer checks and next-evidence suggestions are authored assumptions; historical fields are reproduced exactly. Launch calculations use supplied fixtures and state their sequencing assumptions.
Acceptance checks and observed author tests
- Choose each of the five scenarios: the source ID and all five historical fields change together.
- Try “equipment delivery”, “absent volunteer”, “quality rework”, “stakeholder approval” and “demand surge”: each retrieves its matching supplied case.
- Try “supplier and approval”: both candidates are offered, with no arbitrary winner. Try “urgent”: no candidate is invented.
- Simulate uncertain, different, same-with-missing-checks, and same-with-both-checks: each produces a different, explicit disposition.
- Change an assumption after simulating: the old result is withdrawn until a fresh simulation. Reset restores the initial state.