Run a narrow Prototype Probe and fold findings into a draft spec.
npx skills add https://github.com/MajesteitBart/delano --skill prototype-skill
Please help me install this Agent Skill. Skill: prototype-skill Source: https://github.com/MajesteitBart/delano/tree/main/.agents/skills/prototype-skill Command: npx skills add https://github.com/MajesteitBart/delano --skill prototype-skill
This Skill helps teams resolve material uncertainty before approving a spec by running a focused Prototype Probe. It prevents speculative decisions when a delivery, integration, or UX question needs evidence first.
Ask the AI to run a narrow Prototype Probe for the draft spec, bound the specific uncertainty, and return the updated spec with findings and an approval recommendation.
High-intent search queries and answers about installing and using this skill.
To retire delivery risk before spec approval, you run a narrow, time-boxed Prototype Probe that tests a concrete unknown and fold the evidence back into your draft spec to generate a clear approve or revise recommendation.
A prototype probe for spec review is a focused, time-boxed experiment that tests a specific technical, UX, or integration uncertainty to bound delivery risk instead of relying on speculative debate.
You need to run a time-boxed experiment on a draft specification when material uncertainty exists around an unverified integration path or UX question that requires evidence before planning can safely continue.
You convert probe findings into an approval recommendation by documenting touched surfaces and time-boxed evidence, then returning an explicit decision to approve the spec, revise it, or run another narrow probe.
Spec probing works for bounding integration and UX risk by implementing the smallest viable test path, keeping the experiment narrow enough to finish quickly, and converting findings into a clear approval recommendation.
The limitation of running a prototype probe is that the experiment must remain strictly time-boxed and narrow; it cannot resolve broad architectural uncertainties and only yields an explicit approve, revise, or probe-again recommendation.
softwaresaltRun time-boxed technical investigations and produce evidence-based findings artifacts.
vh2224Execute disposable probes to test milestone viability hypotheses.
product-on-purposeDocument time-boxed spike results with findings and recommendations.
pvliesdonkFacilitate multi-agent architectural deliberation among heterogeneous LLMs via GitHub Discussions.
