consult · diff
git:20260817.f04083f to git:20260927.ef2184c
14 added, 9 removed. Audit A to A.
---
name: consult
description: Bring how named experts would think about the question. Use when the user wants expert thinking on a tradeoff, choice, or stuck problem — not a verdict.
---
## Pick
- Read `profiles/` and match the question's domain. Take two or three, preferring positions that pull against each other — experts who would reach different answers here, not different words for the same one.
+ Take the experts the user names. Where they name none, or the question needs an angle theirs miss, read `profiles/` and add one or two whose work bears on it — each bringing a different idea to try, not a side in an argument.
- No profile fits the domain → use experts the model knows well, name them, and say the profile is missing.
+ ## Ideas
- ## Voice
+ Three to five ideas across the panel. The reply opens on idea 1 — no preamble, no method, no note on sources or counts:
- Each expert's thinking goes into the user's head one at a time. That is **explain**, invoked by name once per expert — the settled thing is that expert's documented position, applied to this question.
+ ```
+ 1. <Expert> — **<the change>**
+ *<why, in their voice>*
+ ```
- - Ground every take in the profile or the expert's documented work, reasoning in their own voice — "Fowler argues…", "Hickey would push back…" — carrying the reasoning itself, not a verdict label.
- - Where they disagree, name the disagreement and both sides, and leave it standing. Easy agreement means the picking was wrong — swap one and voice it.
- - The user drives from there. Each pull is deeper on one expert, a different perspective, or who would push back on this. Answer the pull, then wait for the next.
+ The change alone is bold: it is what the user acts on. The name stays plain, the why italic.
- The loop ends when the user says they are done, or starts deciding. Where it ended because they started deciding, the consulting is over at that moment: what the experts said stands as said, and the decision is theirs alone.
+ - The change is concrete: what to cut, add, or swap in the thing in question.
+ - The why is one clause: what that expert would see in this thing, judged on the matter in hand. Their way of thinking, applied — never a record of what they did, a count from their catalogue, or a title of their work.
+ - No line past 72 characters; a line that won't fit is saying two things.
+ - A number appears only where the idea is the number.
+ - Where the thing already exists, each idea is an A/B against it. One line after the ideas says how a pick wins, as `Win: <one clause>`. An idea judged another way carries its own `win:` line under its why, never in the shared one.
## Result
- The experts' reasoning is the result, and it lives in the user's head. No recap, no closing recommendation, no verdict assembled out of what they said. Every disagreement between them stays open. The user decides.
+ The ideas, then the question alone on its last line: `Build all <N>?` Nothing else — no per-expert sections, no disagreement, no recap, no recommendation. Choosing is the user's.
The question has one retrievable answer → retrieve it; stating a fact needs no expert. The user wants a verdict on something that exists → use **judge** skill. The user wants a proposal they already committed to attacked → use **interrogate** skill.