mirror of
https://github.com/pbakaus/impeccable.git
synced 2026-09-12 14:16:28 +03:00
Paul BakausandClaude Fable 5
c1953a379d
Contract-probe campaign findings (evals repo, notes/fable-oneshot-craft-plan.md): a single model's resonance ranking is deterministic (30/35 identical concepts across 16 framings); dice must come from the script, mirroring the palette-seed result. Derived candidates stay grounded in the audience's world + subject's cultural home; challengers win only on identification x clarity; incumbent-with-deliberate-idea overrides the roll. Validated at contract level on 01-observability + r10-lektor (teletext ranks #3 for lektor; assigned index 3 produced it). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
84 lines
3.4 KiB
JavaScript
84 lines
3.4 KiB
JavaScript
#!/usr/bin/env node
|
|
/**
|
|
* Concept-seed picker: the dice half of the new-work concept procedure.
|
|
*
|
|
* The model derives a grounded shortlist of candidate FORMS from the
|
|
* audience's world and the subject's cultural home (see
|
|
* reference/new-work.md). Left alone, it then always builds its #1 —
|
|
* and a single model's resonance ranking is deterministic, so every run
|
|
* in a category ships the same one or two concepts. Measured: 30/35
|
|
* identical concepts across 16 prompt framings; the model cannot roll
|
|
* its own dice.
|
|
*
|
|
* This script rolls them from outside, the same trick that made the
|
|
* palette seed work:
|
|
* - BUILD INDEX (2-5): which entry of the model's own resonance-ordered
|
|
* shortlist to build. The dice never choose an ungrounded ingredient;
|
|
* they only refuse the argmax rut. (Index 1 is excluded: that's the
|
|
* concept every run would ship anyway.)
|
|
* - CHALLENGERS (3): outside forms from concept-ingredients.json, weighed
|
|
* against the derived candidates on exactly two axes — audience
|
|
* identification and product clarity. They win only when they beat the
|
|
* grounded list; measured behavior is that they lose to strong cultural
|
|
* material and win over thin categories, which is the intended shape.
|
|
*
|
|
* Usage:
|
|
* node scripts/concept-seed.mjs # roll at random
|
|
* node scripts/concept-seed.mjs --from <key> # deterministic (hash key)
|
|
*
|
|
* Env vars:
|
|
* IMPECCABLE_CONCEPT_SEED — same as --from; for reproducible eval runs.
|
|
*/
|
|
|
|
import crypto from 'node:crypto';
|
|
import { readFileSync } from 'node:fs';
|
|
import { dirname, join } from 'node:path';
|
|
import { fileURLToPath } from 'node:url';
|
|
|
|
const here = dirname(fileURLToPath(import.meta.url));
|
|
const pool = JSON.parse(readFileSync(join(here, 'concept-ingredients.json'), 'utf8'));
|
|
|
|
const args = process.argv.slice(2);
|
|
const fromIdx = args.indexOf('--from');
|
|
const key = fromIdx !== -1 ? args[fromIdx + 1] : (process.env.IMPECCABLE_CONCEPT_SEED || null);
|
|
|
|
function hashUnit(k, salt) {
|
|
const h = crypto.createHash('sha256').update(`${salt}:${k}`).digest();
|
|
return h.readUInt32BE(0) / 0xffffffff;
|
|
}
|
|
const unit = (salt) => (key ? hashUnit(key, salt) : Math.random());
|
|
|
|
const buildIndex = 2 + Math.floor(unit('index') * 4); // 2..5
|
|
|
|
const entries = Object.entries(pool)
|
|
.filter(([k]) => !k.startsWith('_'))
|
|
.flatMap(([, list]) => list);
|
|
const picks = [];
|
|
const taken = new Set();
|
|
for (let i = 0; picks.length < 3 && i < 60; i++) {
|
|
const idx = Math.floor(unit(`challenger-${i}`) * entries.length) % entries.length;
|
|
if (!taken.has(idx)) {
|
|
taken.add(idx);
|
|
picks.push(entries[idx]);
|
|
}
|
|
}
|
|
|
|
process.stdout.write(`CONCEPT SEED
|
|
BUILD INDEX: ${buildIndex}
|
|
After ordering your derived candidates by resonance, build the page whose
|
|
form comes from candidate number ${buildIndex}, exactly as if it had ranked
|
|
first: full commitment. Your top-ranked candidate is what every run in this
|
|
category would ship; the assignment exists to refuse that rut, not to
|
|
punish it.
|
|
CHALLENGERS (weigh against your derived candidates on the same two axes,
|
|
audience identification and product clarity; a challenger wins only when
|
|
it beats the grounded list on both):
|
|
1. ${picks[0]}
|
|
2. ${picks[1]}
|
|
3. ${picks[2]}
|
|
If a challenger wins, it replaces the assigned candidate. If the surface is
|
|
an existing world whose incumbent carries a deliberate, ownable idea, the
|
|
incumbent IS the chosen candidate: intensify its lineage and ignore the
|
|
roll entirely.
|
|
`);
|