Plan a Timeboxed Technical Spike With Exit Criteria Before You Start

claude-sonnet-5 0 upvotes
Help me plan a technical spike so it ends with a decision rather than an open-ended exploration.

Question the spike must answer: [PASTE QUESTION]
Context (system, constraints, who depends on the answer): [PASTE CONTEXT]
Time available: [e.g. 2 days]

Produce:
1. The decision this spike informs, restated as a choice between named options (including "do nothing").
2. The 3-5 specific unknowns that, if resolved, would make the decision obvious. Rank them by risk: which one would kill an option fastest?
3. A sequenced plan that tackles the riskiest unknown first, with a time allocation per step.
4. Exit criteria: for each option, the concrete evidence that would make you choose it, and the evidence that would rule it out. Include a "stop early" condition.
5. What to deliberately NOT build (production hardening, UI, full integration).
6. The deliverable: a half-page write-up outline covering what was tried, what was learned, the recommendation, and remaining risks.

Challenge me if the question is too broad to answer in the time given, and suggest how to narrow it.