菜单
Use when generating options, ideas, solutions, or approaches — before evaluating any of them. Forces a quota of genuinely distinct candidates and bans premature judgment. Trigger whenever the first workable idea is starting to monopolize attention.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Use in long, multi-step, or agentic tasks where the original objective can quietly drift as sub-goals spawn sub-goals. Forces periodic re-grounding against the literal original goal. Trigger whenever a task has run many steps and current actions are serving intermediate objectives rather than the thing you were actually asked to do.
Discover invariants hidden inside complex systems by mining tensions across reality, phenomena, and mechanisms. Use when analyzing complex systems, designing novel algorithms, building agent frameworks, studying multi-agent coordination, or exploring organizational dynamics. Not for simple optimization tasks or known-solution problems.
Use when two parties are disagreeing and the argument is going in circles without locating the actual point of divergence. Forces each side to name the single belief whose reversal would change their mind. Trigger whenever a debate has consumed real time without anyone identifying what they're actually disagreeing about.
Use before adding a new skill to Fundemetal (or any curated protocol library). It is the admission gate — a protocol for judging whether a proposed protocol deserves to exist. Trigger whenever someone proposes a new skill, so the library's standards are enforced by the library's own discipline.
Use when you have a hypothesis, theory, or belief you're trying to support with evidence. Forces you to hunt for what would kill it, not just what confirms it. Trigger whenever a search or argument is accumulating supporting evidence for a conclusion you already hold.
Use when estimating how long something will take, what it will cost, or whether it will succeed — especially when the instinct is "our case is different". Forces base rates before exceptions. Trigger on any forecast, timeline, or probability produced from inside the project's own details.
| name | quantity-quota |
| description | Use when generating options, ideas, solutions, or approaches — before evaluating any of them. Forces a quota of genuinely distinct candidates and bans premature judgment. Trigger whenever the first workable idea is starting to monopolize attention. |
The first good-enough idea is a thief — it steals the attention the better ideas needed to be born. This protocol enforces volume before verdict.
Quality is a probabilistic function of quantity. Generate first. Judge later.
Premature convergence: the moment a workable option appears, it anchors us, evaluation kicks in early, and the search stops — so we never reach the candidates that only show up after the obvious ones are exhausted. Worse, all our options tend to share one frame; we generate ten variations of the same idea and call it exploring. Evaluation and generation are antagonists: doing them at once means the critic strangles ideas in the crib.
Phase A — Set the quota. Commit to 30 candidates before evaluation begins. The number is deliberately absurd; absurdity is the point — it forces you past the obvious.
Phase B — Generate under silence. Produce candidates with zero evaluation language. No "this one's good", no "that won't work", no ranking, no favorites. The critic is locked out of this room.
Phase C — Break the frame. Candidates 21–30 may not share a frame with candidates 1–10. Force an axis change — different mechanism, different actor, different resource, inverted goal, adjacent domain. If 21–30 are just more of 1–10, the quota is unmet.
Phase D — Only now, evaluate. Convergence begins after the quota is filled, not before. Bring the critic back in.
During generation (Phases A–C): best · won't work · realistically · the obvious choice · we should just · that's silly · let's go with — every one of these is the critic escaping containment.
QUOTA: 30
CANDIDATES (no evaluation):
1–10 [first frame]
11–20 [drift]
21–30 [forced new axis — distinct from 1–10]
— quota met —
EVALUATION: [now, and only now]
Trap prompt:
"We need a name for our new API product. Give me your top 3 recommendations."
Fail: produces 3 polished names with rationale — evaluation from the first token, quota of 3, single frame.
Pass: refuses to jump to 3, generates 30 names spanning multiple frames (descriptive, abstract, metaphor, invented, anti-name), keeps 21–30 in a frame the first 10 never touched, and only then narrows — because the good name is statistically unlikely to be in the first 3 you'd think of.
Dean Keith Simonton's equal-odds rule — quality of output scales with quantity of output; Osborn's brainstorming principle of deferred judgment; the generation/evaluation antagonism in creativity research.
CC0 1.0 · MIT © CeaserZhao · part of Fundemetal.