LuckyMouse420

Untitled

Sep 6th, 2026
75
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 37.93 KB | None | 0 0
  1. UNIVERSAL RANDOM CHARACTER PROMPT GENERATOR
  2.  
  3. VERSION 1.0 — ZERO-INPUT PARENT GENERATOR + BRANCH/LIMB/LEAF ENTROPY ARCHITECTURE + MODULAR DAUGHTER-PROMPT CONSTRUCTION + SOURCE-AGNOSTIC EXECUTION CONTRACT + STABLE CROSS-REFERENCE BEHAVIOR
  4.  
  5. ══════════════════════════════════════════════════
  6. PURPOSE
  7. ══════════════════════════════════════════════════
  8.  
  9. Generate ONE complete daughter prompt for later use with a character reference image.
  10.  
  11. THIS PARENT GENERATOR RECEIVES NO CHARACTER REFERENCE.
  12.  
  13. THIS PARENT GENERATOR RECEIVES NO CHARACTER DESCRIPTION.
  14.  
  15. THIS PARENT GENERATOR RECEIVES NO DESTINATION REQUEST.
  16.  
  17. Its job is to independently decide what kind of character-conversion prompt to create.
  18.  
  19. The parent generator chooses:
  20.  
  21. WHAT KIND OF TRANSFORMATION OR REINTERPRETATION
  22. THE FUTURE DAUGHTER PROMPT WILL PERFORM.
  23.  
  24. The daughter prompt later receives:
  25.  
  26. @Image1 = Character Reference
  27.  
  28. and applies the already-selected concept to whatever character is attached.
  29.  
  30. Use:
  31.  
  32. NO SOURCE
  33. →
  34. RANDOM BRANCH
  35. →
  36. RANDOM LIMB
  37. →
  38. OPEN LEAF GENERATION
  39. →
  40. RANDOM LEAF
  41. →
  42. RANDOM SUPPORTING VARIABLES
  43. →
  44. DAUGHTER PROMPT COMPILER
  45. →
  46. ONE COMPLETE UNIVERSAL PROMPT
  47. →
  48. STOP.
  49.  
  50. The parent chooses the concept.
  51.  
  52. The daughter applies the concept.
  53.  
  54. The future character reference supplies identity only.
  55.  
  56. ══════════════════════════════════════════════════
  57. CORE CONTRACT
  58. ══════════════════════════════════════════════════
  59.  
  60. INPUT:
  61.  
  62. NONE.
  63.  
  64. OUTPUT:
  65.  
  66. ONE complete daughter prompt.
  67.  
  68. The daughter prompt must be usable later with ANY suitable character reference.
  69.  
  70. The daughter prompt must not contain source-specific assumptions.
  71.  
  72. It must not assume:
  73.  
  74. - species;
  75. - gender;
  76. - body plan;
  77. - age beyond explicit safety requirements;
  78. - occupation;
  79. - faction;
  80. - setting;
  81. - powers;
  82. - clothing;
  83. - personality;
  84. - anatomy;
  85. - world;
  86. - art style;
  87. - source presentation format.
  88.  
  89. The daughter prompt must produce the SAME governing concept regardless of which character reference is later attached.
  90.  
  91. Different references should yield:
  92.  
  93. THE SAME CONCEPT
  94. +
  95. DIFFERENT CHARACTER IDENTITY.
  96.  
  97. ══════════════════════════════════════════════════
  98. NO-SOURCE ARCHITECTURE
  99. ══════════════════════════════════════════════════
  100.  
  101. The parent generator must operate without examining or imagining a future reference.
  102.  
  103. Do not ask:
  104.  
  105. - who the character is;
  106. - what species they are;
  107. - what they look like;
  108. - what would suit them;
  109. - what would contrast with them;
  110. - what would be funny for them;
  111. - what would preserve their lore;
  112. - what would be aesthetically compatible.
  113.  
  114. There is no source during parent generation.
  115.  
  116. Therefore:
  117.  
  118. SOURCE BIAS IS STRUCTURALLY IMPOSSIBLE.
  119.  
  120. The parent must not simulate a hypothetical source either.
  121.  
  122. Do not secretly assume:
  123.  
  124. - humanoid;
  125. - fantasy character;
  126. - anime character;
  127. - woman;
  128. - man;
  129. - animal;
  130. - monster;
  131. - superhero;
  132. - military character.
  133.  
  134. Generate the daughter prompt first.
  135.  
  136. A future reference will adapt to it later.
  137.  
  138. ══════════════════════════════════════════════════
  139. ENTROPY ARCHITECTURE
  140. ══════════════════════════════════════════════════
  141.  
  142. Use independent opaque numeric entropy tokens.
  143.  
  144. Use integers:
  145.  
  146. 0000–9999
  147.  
  148. for all entropy events.
  149.  
  150. Treat the numbers as semantically meaningless.
  151.  
  152. The number determines the result.
  153.  
  154. The result must never determine the number.
  155.  
  156. Use the following hierarchy:
  157.  
  158. E1
  159. →
  160. BRANCH
  161.  
  162. E2
  163. →
  164. LIMB
  165.  
  166. LEAF GENERATION
  167. →
  168. HIDDEN CANDIDATE POOL
  169.  
  170. E3
  171. →
  172. FINAL LEAF
  173.  
  174. Then optionally:
  175.  
  176. E4
  177. →
  178. DAUGHTER PROMPT COMPLEXITY
  179.  
  180. E5
  181. →
  182. ACTIVITY / STATE MODULE
  183.  
  184. E6
  185. →
  186. WORLD / CONTEXT MODULE
  187.  
  188. E7
  189. →
  190. MOMENT / SITUATION MODULE
  191.  
  192. E8
  193. →
  194. PRESENTATION MODULE
  195.  
  196. E9
  197. →
  198. TONE MODULE
  199.  
  200. Additional entropy events may be used only for downstream presentation variables.
  201.  
  202. Do not use additional entropy to reopen or replace the selected core leaf.
  203.  
  204. ══════════════════════════════════════════════════
  205. ENTROPY SOURCE RULE
  206. ══════════════════════════════════════════════════
  207.  
  208. If a genuine external random-number mechanism is available:
  209.  
  210. USE IT.
  211.  
  212. Otherwise instantiate semantically indifferent numeric tokens.
  213.  
  214. Do not choose values because they:
  215.  
  216. - look random;
  217. - create novelty;
  218. - avoid repetition;
  219. - create a better result;
  220. - create a stranger result;
  221. - create an easier daughter prompt;
  222. - correspond to a preferred branch.
  223.  
  224. Use:
  225.  
  226. TOKEN
  227. →
  228. MECHANICAL MAPPING
  229. →
  230. LOCK.
  231.  
  232. ══════════════════════════════════════════════════
  233. E1 — BRANCH
  234. ══════════════════════════════════════════════════
  235.  
  236. Map E1 mechanically:
  237.  
  238. 0000–1999
  239. → ROLE / PERSON-STATE
  240.  
  241. 2000–3999
  242. → LIVING CREATURE / BEING
  243.  
  244. 4000–5999
  245. → MANUFACTURED OBJECT / ARTIFACT / STRUCTURE
  246.  
  247. 6000–6999
  248. → PLANT / FUNGUS / BOTANICAL LIFE
  249.  
  250. 7000–7999
  251. → NATURAL FORMATION / PLACE / CELESTIAL BODY
  252.  
  253. 8000–8999
  254. → PHENOMENON / FORCE / ABSTRACT EMBODIMENT
  255.  
  256. 9000–9999
  257. → UNRESTRICTED WILDCARD
  258.  
  259. After mapping:
  260.  
  261. LOCK THE BRANCH.
  262.  
  263. Do not reroll.
  264.  
  265. ══════════════════════════════════════════════════
  266. E2 — LIMB
  267. ══════════════════════════════════════════════════
  268.  
  269. After branch lock:
  270.  
  271. instantiate E2.
  272.  
  273. Map E2 only within the selected branch.
  274.  
  275. ══════════════════════════════════════════════════
  276. ROLE / PERSON-STATE LIMBS
  277. ══════════════════════════════════════════════════
  278.  
  279. 0000–1799
  280. → occupation / trade / vocation
  281.  
  282. 1800–3199
  283. → sport / athletic / competitive role
  284.  
  285. 3200–4399
  286. → performance / entertainment role
  287.  
  288. 4400–5599
  289. → institutional / civic / service role
  290.  
  291. 5600–6699
  292. → cultural / social / ceremonial role
  293.  
  294. 6700–7699
  295. → leadership / authority / governance role
  296.  
  297. 7700–8599
  298. → historical / era-bound person-state
  299.  
  300. 8600–9499
  301. → situational / temporary person-state
  302.  
  303. 9500–9999
  304. → branch-local wildcard
  305.  
  306. ══════════════════════════════════════════════════
  307. LIVING CREATURE / BEING LIMBS
  308. ══════════════════════════════════════════════════
  309.  
  310. 0000–1399
  311. → terrestrial vertebrate life
  312.  
  313. 1400–2799
  314. → aquatic / marine life
  315.  
  316. 2800–3799
  317. → aerial / flight-specialized life
  318.  
  319. 3800–5299
  320. → invertebrate life
  321.  
  322. 5300–6299
  323. → microscopic / colonial / swarm-scale life
  324.  
  325. 6300–7399
  326. → extinct / prehistoric life
  327.  
  328. 7400–8399
  329. → extraterrestrial organism
  330.  
  331. 8400–9499
  332. → supernatural / monstrous / anomalous being
  333.  
  334. 9500–9999
  335. → branch-local wildcard
  336.  
  337. ══════════════════════════════════════════════════
  338. MANUFACTURED OBJECT / ARTIFACT / STRUCTURE LIMBS
  339. ══════════════════════════════════════════════════
  340.  
  341. 0000–1399
  342. → domestic object
  343.  
  344. 1400–2599
  345. → civic / public object
  346.  
  347. 2600–3999
  348. → tool / instrument / device
  349.  
  350. 4000–5099
  351. → furniture / fixture
  352.  
  353. 5100–6199
  354. → vehicle / transport object
  355.  
  356. 6200–7299
  357. → machine / appliance
  358.  
  359. 7300–8399
  360. → architecture / built structure
  361.  
  362. 8400–9499
  363. → infrastructure / utility system
  364.  
  365. 9500–9999
  366. → branch-local wildcard
  367.  
  368. ══════════════════════════════════════════════════
  369. PLANT / FUNGUS / BOTANICAL LIFE LIMBS
  370. ══════════════════════════════════════════════════
  371.  
  372. 0000–1799
  373. → tree / woody growth
  374.  
  375. 1800–3199
  376. → herbaceous / leafy plant
  377.  
  378. 3200–4599
  379. → flowering / fruiting plant
  380.  
  381. 4600–5699
  382. → vine / climbing / creeping growth
  383.  
  384. 5700–6799
  385. → grass / reed / groundcover form
  386.  
  387. 6800–8399
  388. → fungus / mushroom / mycelial form
  389.  
  390. 8400–9499
  391. → aquatic / unusual botanical form
  392.  
  393. 9500–9999
  394. → branch-local wildcard
  395.  
  396. ══════════════════════════════════════════════════
  397. NATURAL FORMATION / PLACE / CELESTIAL BODY LIMBS
  398. ══════════════════════════════════════════════════
  399.  
  400. 0000–1899
  401. → geological / landform
  402.  
  403. 1900–3299
  404. → hydrological / water formation
  405.  
  406. 3300–4299
  407. → frozen / cryogenic natural form
  408.  
  409. 4300–5499
  410. → ecosystem / habitat-scale formation
  411.  
  412. 5500–6699
  413. → geographic place / terrain region
  414.  
  415. 6700–7999
  416. → planetary / lunar / minor celestial body
  417.  
  418. 8000–9499
  419. → stellar / deep-space / galactic structure
  420.  
  421. 9500–9999
  422. → branch-local wildcard
  423.  
  424. ══════════════════════════════════════════════════
  425. PHENOMENON / FORCE / ABSTRACT EMBODIMENT LIMBS
  426. ══════════════════════════════════════════════════
  427.  
  428. 0000–1399
  429. → atmospheric / weather phenomenon
  430.  
  431. 1400–2499
  432. → fluid / wave / turbulence phenomenon
  433.  
  434. 2500–3599
  435. → thermal / energetic process
  436.  
  437. 3600–4699
  438. → field / force / pressure / attraction
  439.  
  440. 4700–5799
  441. → optical / acoustic / signal phenomenon
  442.  
  443. 5800–6799
  444. → temporal / cyclical / persistence phenomenon
  445.  
  446. 6800–7799
  447. → informational / mathematical / relational structure
  448.  
  449. 7800–8699
  450. → psychological / symbolic / conceptual embodiment
  451.  
  452. 8700–9499
  453. → anomalous / impossible process
  454.  
  455. 9500–9999
  456. → branch-local wildcard
  457.  
  458. ══════════════════════════════════════════════════
  459. UNRESTRICTED WILDCARD
  460. ══════════════════════════════════════════════════
  461.  
  462. If E1 selects unrestricted wildcard:
  463.  
  464. E2 does not constrain the search to a normal branch.
  465.  
  466. Use E2 only as additional entropy.
  467.  
  468. Generate leaves from the complete conceptual possibility space.
  469.  
  470. Wildcard does NOT mean:
  471.  
  472. - deliberately bizarre;
  473. - deliberately surreal;
  474. - deliberately spectacular;
  475. - deliberately abstract.
  476.  
  477. It means:
  478.  
  479. NO ONTOLOGICAL FAMILY RESTRICTION.
  480.  
  481. ══════════════════════════════════════════════════
  482. OPEN LEAF GENERATION
  483. ══════════════════════════════════════════════════
  484.  
  485. After branch and limb lock:
  486.  
  487. generate a hidden pool of concrete daughter-prompt premises.
  488.  
  489. The pool must remain OPEN-ENDED.
  490.  
  491. Do not use a fixed endpoint catalog.
  492.  
  493. Do not rely on concrete examples from this parent prompt.
  494.  
  495. Do not merely recall the most famous or obvious member of the selected limb.
  496.  
  497. Generate multiple substantially distinct concrete leaves from the selected conceptual territory.
  498.  
  499. A leaf must define one clear governing conversion premise.
  500.  
  501. Possible leaf scale may range naturally from:
  502.  
  503. simple
  504. to
  505. complex.
  506.  
  507. Do not prefer complexity.
  508.  
  509. Do not prefer novelty.
  510.  
  511. Do not prefer spectacle.
  512.  
  513. Do not prefer humor.
  514.  
  515. Do not prefer ordinary results.
  516.  
  517. ══════════════════════════════════════════════════
  518. LEAF-POOL DIVERSITY AUDIT
  519. ══════════════════════════════════════════════════
  520.  
  521. Before E3 exists:
  522.  
  523. inspect the hidden leaf pool.
  524.  
  525. Check for:
  526.  
  527. - near-duplicate concepts;
  528. - multiple cosmetic variants of one archetype;
  529. - repeated silhouettes;
  530. - repeated narrative functions;
  531. - repeated species neighborhoods;
  532. - repeated occupational neighborhoods;
  533. - repeated structural archetypes;
  534. - repeated famous exemplars;
  535. - repeated transformation mechanics.
  536.  
  537. If several candidates are substantially the same governing concept:
  538.  
  539. replace enough candidates to restore meaningful spread.
  540.  
  541. Do not optimize for obscurity.
  542.  
  543. A familiar leaf remains valid.
  544.  
  545. The goal is:
  546.  
  547. NO UNJUSTIFIED CONCEPTUAL DENSITY.
  548.  
  549. ══════════════════════════════════════════════════
  550. POOL FREEZE
  551. ══════════════════════════════════════════════════
  552.  
  553. After diversity audit:
  554.  
  555. FREEZE:
  556.  
  557. - pool membership;
  558. - candidate order.
  559.  
  560. Only then instantiate E3.
  561.  
  562. ══════════════════════════════════════════════════
  563. E3 — FINAL LEAF
  564. ══════════════════════════════════════════════════
  565.  
  566. For N frozen candidates:
  567.  
  568. INDEX
  569. =
  570. floor((E3 / 10000) × N)
  571.  
  572. using zero-based indexing.
  573.  
  574. The mapped leaf is binding.
  575.  
  576. Do not:
  577.  
  578. - reroll;
  579. - reorder;
  580. - replace;
  581. - optimize;
  582. - soften;
  583. - escalate;
  584. - simplify;
  585. - choose a neighboring concept.
  586.  
  587. LOCK THE LEAF.
  588.  
  589. This leaf defines the central purpose of the daughter prompt.
  590.  
  591. ══════════════════════════════════════════════════
  592. CORE LEAF ASSERTION
  593. ══════════════════════════════════════════════════
  594.  
  595. The selected leaf must be expressible internally as:
  596.  
  597. THE FUTURE CHARACTER WILL BE DEPICTED AS:
  598. [ONE CONCRETE CONCEPT]
  599.  
  600. or:
  601.  
  602. THE FUTURE CHARACTER WILL LITERALLY BECOME:
  603. [ONE CONCRETE DESTINATION]
  604.  
  605. or:
  606.  
  607. THE FUTURE CHARACTER WILL OCCUPY:
  608. [ONE CONCRETE ROLE / STATE / SITUATION]
  609.  
  610. Use the shallowest statement that accurately describes the leaf.
  611.  
  612. This governing assertion may not change downstream.
  613.  
  614. ══════════════════════════════════════════════════
  615. DAUGHTER PROMPT MODE
  616. ══════════════════════════════════════════════════
  617.  
  618. Determine what kind of daughter prompt the leaf requires.
  619.  
  620. Possible structural modes include:
  621.  
  622. IDENTITY-PRESERVING ROLE CONVERSION
  623.  
  624. LITERAL SPECIES / ORGANISM CONVERSION
  625.  
  626. LITERAL OBJECT / STRUCTURE CONVERSION
  627.  
  628. BOTANICAL CONVERSION
  629.  
  630. PLACE / FORMATION / CELESTIAL CONVERSION
  631.  
  632. PHENOMENON / ABSTRACT CONVERSION
  633.  
  634. SCENARIO-CENTERED CHARACTER PROMPT
  635.  
  636. HYBRID MODE
  637. only when the locked leaf itself genuinely requires several coupled elements.
  638.  
  639. Do not use hybrid mode merely to add complexity.
  640.  
  641. ══════════════════════════════════════════════════
  642. E4 — PROMPT COMPLEXITY
  643. ══════════════════════════════════════════════════
  644.  
  645. Instantiate E4 after leaf lock.
  646.  
  647. Map:
  648.  
  649. 0000–2499
  650. → MINIMAL
  651.  
  652. 2500–5999
  653. → MODERATE
  654.  
  655. 6000–8499
  656. → RICH
  657.  
  658. 8500–9999
  659. → ELABORATE
  660.  
  661. Complexity determines how many supporting modules may activate.
  662.  
  663. It does NOT alter the core leaf.
  664.  
  665. MINIMAL:
  666.  
  667. use only essential execution information.
  668.  
  669. MODERATE:
  670.  
  671. allow a limited scenario and presentation layer.
  672.  
  673. RICH:
  674.  
  675. allow several coherent supporting variables.
  676.  
  677. ELABORATE:
  678.  
  679. allow a strongly developed scenario with detailed presentation.
  680.  
  681. Do not confuse elaborate with spectacular.
  682.  
  683. An elaborate prompt may describe an ordinary situation in detail.
  684.  
  685. ══════════════════════════════════════════════════
  686. OPTIONAL SUPPORT MODULES
  687. ══════════════════════════════════════════════════
  688.  
  689. Supporting modules may enrich the daughter prompt.
  690.  
  691. They may never replace the leaf.
  692.  
  693. Use independent entropy events.
  694.  
  695. ══════════════════════════════════════════════════
  696. E5 — ACTIVITY / STATE
  697. ══════════════════════════════════════════════════
  698.  
  699. Determine activation mechanically based on E5 and complexity.
  700.  
  701. If active:
  702.  
  703. select ONE activity, behavior, task, state, performance, interaction, or occupation that is logically compatible with the leaf.
  704.  
  705. If the leaf already specifies this:
  706.  
  707. do not redundantly add another.
  708.  
  709. The activity must answer:
  710.  
  711. WHAT IS THE SUBJECT DOING?
  712.  
  713. The activity is subordinate to the core destination.
  714.  
  715. ══════════════════════════════════════════════════
  716. E6 — WORLD / CONTEXT
  717. ══════════════════════════════════════════════════
  718.  
  719. If active:
  720.  
  721. select ONE contextual environment or world condition.
  722.  
  723. The context may establish:
  724.  
  725. - era;
  726. - location type;
  727. - cultural frame;
  728. - social setting;
  729. - environmental conditions;
  730. - speculative world logic;
  731. - ordinary real-world context.
  732.  
  733. Do not default to:
  734.  
  735. - fantasy;
  736. - science fiction;
  737. - cyberpunk;
  738. - royal;
  739. - gothic;
  740. - magical;
  741. - post-apocalyptic;
  742. - cinematic spectacle.
  743.  
  744. Context must follow the selected leaf and remain secondary to it.
  745.  
  746. ══════════════════════════════════════════════════
  747. E7 — MOMENT / SITUATION
  748. ══════════════════════════════════════════════════
  749.  
  750. If active:
  751.  
  752. select ONE concrete moment.
  753.  
  754. Answer:
  755.  
  756. WHAT IS HAPPENING RIGHT NOW?
  757.  
  758. The moment may be:
  759.  
  760. - routine;
  761. - transitional;
  762. - active;
  763. - social;
  764. - solitary;
  765. - competitive;
  766. - ceremonial;
  767. - awkward;
  768. - quiet;
  769. - chaotic;
  770. - consequential;
  771. - mundane.
  772.  
  773. Do not automatically choose:
  774.  
  775. - battle;
  776. - confrontation;
  777. - victory pose;
  778. - dramatic reveal.
  779.  
  780. ══════════════════════════════════════════════════
  781. E8 — PRESENTATION MODE
  782. ══════════════════════════════════════════════════
  783.  
  784. Select one presentation strategy appropriate to the final concept.
  785.  
  786. The presentation may govern:
  787.  
  788. - framing;
  789. - camera distance;
  790. - composition;
  791. - image language;
  792. - scene density;
  793. - degree of environmental storytelling.
  794.  
  795. Possible broad territories include:
  796.  
  797. portrait-oriented
  798. scene-oriented
  799. action-oriented
  800. documentary-oriented
  801. editorial-oriented
  802. illustrative
  803. photographic
  804. cinematic
  805. candid
  806. formal
  807. diagrammatic only when genuinely appropriate
  808. unrestricted wildcard
  809.  
  810. Do not reproduce a future reference image's layout automatically.
  811.  
  812. The daughter prompt must explicitly treat the eventual source as identity authority, not composition authority.
  813.  
  814. ══════════════════════════════════════════════════
  815. E9 — TONE
  816. ══════════════════════════════════════════════════
  817.  
  818. Select one tonal direction.
  819.  
  820. Tone affects atmosphere only.
  821.  
  822. It may not redefine the leaf.
  823.  
  824. Possible broad territories include:
  825.  
  826. restrained
  827. serious
  828. playful
  829. energetic
  830. peaceful
  831. awkward
  832. eerie
  833. elegant
  834. melancholic
  835. comedic
  836. chaotic
  837. celebratory
  838. tense
  839. ordinary
  840. unrestricted wildcard
  841.  
  842. Do not assume unusual concepts require dramatic tone.
  843.  
  844. ══════════════════════════════════════════════════
  845. COMPATIBILITY COMPILER
  846. ══════════════════════════════════════════════════
  847.  
  848. After all random variables are locked:
  849.  
  850. compile them into one coherent daughter prompt.
  851.  
  852. The compiler is NOT another randomizer.
  853.  
  854. It may:
  855.  
  856. - resolve contradiction;
  857. - remove redundant modules;
  858. - choose grammatically coherent wording;
  859. - clarify physical execution;
  860. - clarify destination-native anatomy;
  861. - clarify clothing compatibility;
  862. - clarify environment;
  863. - clarify scale;
  864. - clarify subject singularity;
  865. - clarify camera framing;
  866. - clarify source-trait translation.
  867.  
  868. It may NOT:
  869.  
  870. - replace the leaf;
  871. - reroll;
  872. - create a neighboring premise;
  873. - add an unrelated second concept;
  874. - change the result because it is difficult;
  875. - improve the concept by making it more dramatic;
  876. - weaken literal transformation into costume or theme.
  877.  
  878. Use:
  879.  
  880. LOCKED RANDOM VARIABLES
  881. →
  882. NONCREATIVE COMPILATION
  883. →
  884. DAUGHTER PROMPT.
  885.  
  886. ══════════════════════════════════════════════════
  887. DAUGHTER PROMPT SOURCE CONTRACT
  888. ══════════════════════════════════════════════════
  889.  
  890. Every generated daughter prompt must begin from the assumption:
  891.  
  892. @Image1 = Character Reference
  893.  
  894. The future source image establishes WHO the subject is.
  895.  
  896. The daughter prompt's locked concept establishes WHAT is being depicted.
  897.  
  898. The daughter prompt must instruct the future image generator to preserve recognizable source identity while obeying the locked concept.
  899.  
  900. Do not encode specific traits that have not yet been seen.
  901.  
  902. Use generic identity language.
  903.  
  904. For example, refer abstractly to:
  905.  
  906. - recognizable face;
  907. - hairstyle;
  908. - coloration;
  909. - species traits where compatible;
  910. - markings;
  911. - signature accessories where compatible;
  912. - proportions;
  913. - recurring motifs;
  914. - personality expression where useful.
  915.  
  916. Do not invent actual traits.
  917.  
  918. ══════════════════════════════════════════════════
  919. STABLE CROSS-REFERENCE BEHAVIOR
  920. ══════════════════════════════════════════════════
  921.  
  922. The daughter prompt must produce the same governing interpretation when applied to different references.
  923.  
  924. If the same daughter prompt is paired separately with:
  925.  
  926. REFERENCE A
  927. REFERENCE B
  928. REFERENCE C
  929.  
  930. then all three outputs should share:
  931.  
  932. - the same core destination;
  933. - the same role or situation where specified;
  934. - the same setting where specified;
  935. - the same activity where specified;
  936. - the same presentation logic where specified.
  937.  
  938. They should differ only because the attached character identities differ.
  939.  
  940. Use:
  941.  
  942. SAME DAUGHTER PROMPT
  943. +
  944. DIFFERENT REFERENCES
  945. =
  946. SAME CONCEPT
  947. +
  948. DIFFERENT CHARACTER.
  949.  
  950. ══════════════════════════════════════════════════
  951. DESTINATION LITERALITY
  952. ══════════════════════════════════════════════════
  953.  
  954. When the locked leaf defines an ontological conversion:
  955.  
  956. the daughter prompt must make the conversion literal.
  957.  
  958. The image generator must not satisfy the prompt by leaving the source unchanged and adding:
  959.  
  960. - costume;
  961. - themed clothing;
  962. - props;
  963. - background motifs;
  964. - symbolic effects;
  965. - decorative anatomy;
  966. - an adjacent entity;
  967. - a separate transformed object.
  968.  
  969. The referenced character themselves must genuinely occupy the selected ontology.
  970.  
  971. If the destination is naturally nonhumanoid:
  972.  
  973. do not preserve humanoid anatomy merely for recognizability.
  974.  
  975. Translate identity through destination-native structure.
  976.  
  977. ══════════════════════════════════════════════════
  978. ROLE LITERALITY
  979. ══════════════════════════════════════════════════
  980.  
  981. When the locked leaf defines a role or person-state:
  982.  
  983. the daughter prompt must depict the character genuinely occupying that state.
  984.  
  985. Do not merely dress them in suggestive clothing.
  986.  
  987. Show sufficient contextual evidence to establish the role.
  988.  
  989. Avoid turning every role into a glamorous costume portrait.
  990.  
  991. ══════════════════════════════════════════════════
  992. SOURCE-IDENTITY PRESERVATION
  993. ══════════════════════════════════════════════════
  994.  
  995. The daughter prompt must preserve enough of the future reference to keep the transformed subject recognizable.
  996.  
  997. Identity may survive through:
  998.  
  999. - face;
  1000. - head shape;
  1001. - hairstyle;
  1002. - characteristic coloration;
  1003. - markings;
  1004. - signature visual motifs;
  1005. - proportion relationships;
  1006. - distinctive accessories where compatible;
  1007. - translated structural motifs;
  1008. - characteristic expression or attitude.
  1009.  
  1010. Use:
  1011.  
  1012. MAXIMUM CONCEPT FIDELITY
  1013. +
  1014. MINIMUM NECESSARY IDENTITY CONTINUITY.
  1015.  
  1016. Do not allow identity preservation to undo the selected concept.
  1017.  
  1018. ══════════════════════════════════════════════════
  1019. REFERENCE AUTHORITY FIREWALL
  1020. ══════════════════════════════════════════════════
  1021.  
  1022. Every daughter prompt must include an instruction equivalent to:
  1023.  
  1024. Use @Image1 only as authority for the character's identity.
  1025.  
  1026. Do not automatically inherit from the reference:
  1027.  
  1028. - pose;
  1029. - background;
  1030. - panel layout;
  1031. - turnaround structure;
  1032. - expression sheet;
  1033. - labels;
  1034. - text;
  1035. - framing;
  1036. - art-board composition;
  1037. - source-world environment.
  1038.  
  1039. The generated prompt controls presentation.
  1040.  
  1041. ══════════════════════════════════════════════════
  1042. ANTI-FORMAT-BLEED RULE
  1043. ══════════════════════════════════════════════════
  1044.  
  1045. A future reference sheet must not cause the daughter prompt to produce another reference sheet unless the daughter prompt specifically requests one.
  1046.  
  1047. A portrait source must not force a portrait output.
  1048.  
  1049. A comic source must not force a comic output.
  1050.  
  1051. A screenshot must not force screenshot presentation.
  1052.  
  1053. Identity and presentation remain separate.
  1054.  
  1055. ══════════════════════════════════════════════════
  1056. DAUGHTER PROMPT EXECUTION LANGUAGE
  1057. ══════════════════════════════════════════════════
  1058.  
  1059. The final daughter prompt should behave as an authoritative image request.
  1060.  
  1061. It should not tell the future renderer to:
  1062.  
  1063. - randomize;
  1064. - choose;
  1065. - decide;
  1066. - discover;
  1067. - invent a destination;
  1068. - reinterpret freely;
  1069. - select a transformation.
  1070.  
  1071. The destination has already been chosen.
  1072.  
  1073. Use declarative execution language.
  1074.  
  1075. Example structure:
  1076.  
  1077. @Image1 = Character Reference
  1078.  
  1079. Create one finished image depicting the character from @Image1 as [LOCKED CONCEPT].
  1080.  
  1081. [Literal concept requirements.]
  1082.  
  1083. [Optional activity / setting / moment.]
  1084.  
  1085. Preserve the character's recognizable identity from @Image1 while translating incompatible traits into concept-native equivalents.
  1086.  
  1087. [Presentation instructions.]
  1088.  
  1089. [Necessary anti-drift boundaries.]
  1090.  
  1091. No text unless specifically required.
  1092.  
  1093. ══════════════════════════════════════════════════
  1094. NEGATIVE BOUNDARY
  1095. ══════════════════════════════════════════════════
  1096.  
  1097. Add only the negative instructions necessary for the selected leaf.
  1098.  
  1099. Possible categories:
  1100.  
  1101. - not merely a costume;
  1102. - not merely a themed version of the original body;
  1103. - not a separate destination beside the source;
  1104. - not destination as background;
  1105. - not destination as prop;
  1106. - not anthropomorphic fallback;
  1107. - not source-sheet layout;
  1108. - not symbolic substitute.
  1109.  
  1110. Do not dump every negative rule into every daughter prompt.
  1111.  
  1112. Use destination-specific boundaries.
  1113.  
  1114. ══════════════════════════════════════════════════
  1115. DAUGHTER PROMPT QUALITY FLOOR
  1116. ══════════════════════════════════════════════════
  1117.  
  1118. The final daughter prompt must be:
  1119.  
  1120. - self-contained;
  1121. - immediately executable;
  1122. - clear;
  1123. - concrete;
  1124. - source-agnostic;
  1125. - identity-preserving;
  1126. - concept-stable;
  1127. - grammatically complete;
  1128. - free of parent-generator mechanics.
  1129.  
  1130. It must not require the future user to fill in missing creative decisions.
  1131.  
  1132. The only expected external input is:
  1133.  
  1134. @Image1.
  1135.  
  1136. ══════════════════════════════════════════════════
  1137. PARENT OUTPUT RULE
  1138. ══════════════════════════════════════════════════
  1139.  
  1140. The parent generator outputs exactly:
  1141.  
  1142. DAUGHTER PROMPT
  1143.  
  1144. [complete daughter prompt]
  1145.  
  1146. Do not output:
  1147.  
  1148. - entropy values;
  1149. - branch;
  1150. - limb;
  1151. - candidate pool;
  1152. - rejected leaves;
  1153. - probability calculations;
  1154. - audits;
  1155. - alternative prompts;
  1156. - explanation.
  1157.  
  1158. Do not render an image.
  1159.  
  1160. Do not ask for a reference.
  1161.  
  1162. Do not ask a follow-up question.
  1163.  
  1164. Generate one daughter prompt and stop.
  1165.  
  1166. ══════════════════════════════════════════════════
  1167. REROLL BEHAVIOR
  1168. ══════════════════════════════════════════════════
  1169.  
  1170. If the parent generator is run again:
  1171.  
  1172. perform a completely new independent run.
  1173.  
  1174. Do not modify the previous daughter prompt.
  1175.  
  1176. Do not intentionally avoid the previous result.
  1177.  
  1178. Do not intentionally repeat the previous result.
  1179.  
  1180. Previous outputs have no vote.
  1181.  
  1182. A legitimate repeat remains possible.
  1183.  
  1184. ══════════════════════════════════════════════════
  1185. ANTI-ATTRACTOR RULE
  1186. ══════════════════════════════════════════════════
  1187.  
  1188. Do not repeatedly favor concepts because they are:
  1189.  
  1190. - visually impressive;
  1191. - culturally iconic;
  1192. - common in training data;
  1193. - easy to render;
  1194. - strongly associated with transformation art;
  1195. - fantasy-coded;
  1196. - glamorous;
  1197. - monstrous;
  1198. - mechanical;
  1199. - aquatic;
  1200. - celestial;
  1201. - magical;
  1202. - comedic.
  1203.  
  1204. Likewise, do not deliberately suppress them.
  1205.  
  1206. The objective is not forced novelty.
  1207.  
  1208. The objective is:
  1209.  
  1210. NO UNJUSTIFIED SEMANTIC GRAVITY.
  1211.  
  1212. ══════════════════════════════════════════════════
  1213. FAILURE MODES
  1214. ══════════════════════════════════════════════════
  1215.  
  1216. Reject or repair:
  1217.  
  1218. SOURCE ASSUMPTION
  1219. The parent silently imagines what kind of character will later be attached.
  1220.  
  1221. SOURCE-SPECIFIC DAUGHTER PROMPT
  1222. The daughter prompt contains traits that were never provided.
  1223.  
  1224. LEAF OPTIMIZATION
  1225. The selected leaf is replaced by a more appealing concept.
  1226.  
  1227. MODULE OVERRIDE
  1228. A supporting module becomes more important than the core leaf.
  1229.  
  1230. CONCEPT SOUP
  1231. Multiple unrelated random concepts are combined.
  1232.  
  1233. SPECTACLE BIAS
  1234. Random outputs drift toward inherently dramatic concepts.
  1235.  
  1236. FANTASY BIAS
  1237. Outputs repeatedly drift toward fantasy transformations.
  1238.  
  1239. SCI-FI BIAS
  1240. Outputs repeatedly drift toward technological transformations.
  1241.  
  1242. ROLE BIAS
  1243. Outputs repeatedly become professions or costumes.
  1244.  
  1245. HUMANOID BIAS
  1246. Nonhuman conversions retain humanoid anatomy without cause.
  1247.  
  1248. COSTUME FALLBACK
  1249. Ontological conversion becomes themed clothing.
  1250.  
  1251. REFERENCE FORMAT BLEED
  1252. The eventual source controls the daughter's presentation.
  1253.  
  1254. DAUGHTER RANDOMIZATION
  1255. The daughter prompt asks the renderer to decide the concept.
  1256.  
  1257. UNSTABLE CROSS-REFERENCE EXECUTION
  1258. The same daughter prompt produces fundamentally different concepts depending on the attached source.
  1259.  
  1260. OVER-COMPILATION
  1261. The compiler changes the governing leaf.
  1262.  
  1263. UNDER-COMPILATION
  1264. The daughter prompt leaves the renderer to decide what the concept means.
  1265.  
  1266. RANDOM TRAIT SOUP
  1267. Supporting modules create unrelated novelty rather than one coherent image.
  1268.  
  1269. ══════════════════════════════════════════════════
  1270. INTERNAL PARENT CHECK
  1271. ══════════════════════════════════════════════════
  1272.  
  1273. Before output, verify internally:
  1274.  
  1275. 1. Did the parent receive no source information?
  1276. 2. Was E1 instantiated before semantic outcome consideration?
  1277. 3. Was E1 mechanically mapped?
  1278. 4. Was the branch locked?
  1279. 5. Was E2 instantiated only after branch lock?
  1280. 6. Was the limb mechanically mapped?
  1281. 7. Was the limb locked?
  1282. 8. Was the leaf pool generated only after limb lock?
  1283. 9. Was the pool open-ended rather than catalog-based?
  1284. 10. Was conceptual clustering audited?
  1285. 11. Was exemplar density audited?
  1286. 12. Was the pool frozen before E3?
  1287. 13. Was E3 mechanically mapped?
  1288. 14. Was the leaf locked?
  1289. 15. Was there any result-dependent rerolling?
  1290. 16. Did supporting modules remain subordinate?
  1291. 17. Did the compiler preserve the exact leaf?
  1292. 18. Is the daughter prompt source-agnostic?
  1293. 19. Could the daughter prompt accept an arbitrary character reference?
  1294. 20. Would changing the future reference leave the governing concept unchanged?
  1295. 21. Does the daughter prompt contain exactly one core concept?
  1296. 22. Does it contain enough specificity to execute without creative destination inference?
  1297. 23. Did any randomizer language leak into the daughter prompt?
  1298. 24. Is the output one daughter prompt only?
  1299.  
  1300. If any failure exists:
  1301.  
  1302. repair the parent-generation process before output.
  1303.  
  1304. Do not reroll a valid leaf merely to repair wording.
  1305.  
  1306. ══════════════════════════════════════════════════
  1307. MASTER PRINCIPLE
  1308. ══════════════════════════════════════════════════
  1309.  
  1310. DO NOT USE:
  1311.  
  1312. REFERENCE
  1313. →
  1314. RANDOMIZER
  1315. →
  1316. DESTINATION.
  1317.  
  1318. THE PARENT NEVER SEES THE REFERENCE.
  1319.  
  1320. DO NOT USE:
  1321.  
  1322. HYPOTHETICAL CHARACTER
  1323. →
  1324. CHOOSE SUITABLE CONCEPT.
  1325.  
  1326. DO NOT USE:
  1327.  
  1328. RANDOM LEAF
  1329. →
  1330. OPTIMIZE INTO BETTER LEAF.
  1331.  
  1332. DO NOT USE:
  1333.  
  1334. SUPPORT MODULE
  1335. →
  1336. OVERRIDE CORE CONCEPT.
  1337.  
  1338. DO NOT USE:
  1339.  
  1340. DAUGHTER PROMPT
  1341. →
  1342. ASK RENDERER WHAT TO MAKE.
  1343.  
  1344. USE:
  1345.  
  1346. NO INPUT
  1347. →
  1348. E1
  1349. →
  1350. LOCK BRANCH
  1351. →
  1352. E2
  1353. →
  1354. LOCK LIMB
  1355. →
  1356. GENERATE OPEN LEAF POOL
  1357. →
  1358. AUDIT
  1359. →
  1360. FREEZE
  1361. →
  1362. E3
  1363. →
  1364. LOCK LEAF
  1365. →
  1366. ROLL SUPPORT VARIABLES
  1367. →
  1368. COMPILE
  1369. →
  1370. WRITE UNIVERSAL DAUGHTER PROMPT
  1371. →
  1372. STOP.
  1373.  
  1374. LATER:
  1375.  
  1376. DAUGHTER PROMPT
  1377. +
  1378. @Image1
  1379. →
  1380. SAME LOCKED CONCEPT
  1381. +
  1382. REFERENCE-SPECIFIC IDENTITY
  1383. →
  1384. ONE FINISHED IMAGE.
  1385.  
  1386. THE PARENT GENERATOR CHOOSES WHAT KIND OF PROMPT EXISTS.
  1387.  
  1388. THE DAUGHTER PROMPT DEFINES WHAT THE FUTURE IMAGE MUST DEPICT.
  1389.  
  1390. THE FUTURE REFERENCE DEFINES WHOSE IDENTITY IS USED.
  1391.  
  1392. THE IMAGE GENERATOR ONLY EXECUTES THE DAUGHTER PROMPT.
Add Comment
Please, Sign In to add comment