Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- PROMPT MAKER — 4.9 TO CURRENT 5.0 CUMULATIVE UPGRADE PACK
- PACK ID
- PM-4.9-TO-CURRENT-5.0-CUMULATIVE-UPGRADE-PACK
- SOURCE BASELINE
- Prompt Maker Core Architecture 4.9
- TARGET STATE
- Prompt Maker 5.0 architecture PLUS all currently approved 5.0-era upgrades and mutations.
- PURPOSE
- This is the ONE cumulative upgrade file for users who are still on Prompt Maker 4.9
- and want to reach the CURRENT working 5.0 state in one step.
- Use this pack when the goal is:
- 4.9
- → canonical 5.0 baseline
- → all currently approved 5.0-era upgrades
- → one complete upgraded receiver state.
- This pack intentionally combines:
- 1. the 4.9 → 5.0 upgrade layer; and
- 2. the current 5.0 canonical upgrade pack.
- Therefore it includes, in one portable file:
- - the 5.0 baseline migration;
- - the propagation ON/OFF switch behavior;
- - multi-upgrade ingestion;
- - generate upgrade / apply upgrade user-language recognition;
- - baseline-aware upgrade export relative to original 5.0;
- - relationship-first counterpart / villain design logic;
- - semantic-gravity suppression;
- - natural / ecological counterpart testing;
- - conflict-shaped environment reasoning;
- - 2D depth / implied-motion scene logic;
- - source-DNA vs presentation-format firewall;
- - daughter provenance continuity; and
- - daughter-specific capability selection.
- APPLICATION CONTRACT
- Use this as a SEMANTIC UPGRADE PACK.
- Do NOT blindly append it to Prompt Maker 4.9.
- Use:
- FULL 4.9 RECEIVER
- → INTERPRET THIS PACK SEMANTICALLY
- → INTEGRATE THE 5.0 BASELINE CAPABILITIES
- → INTEGRATE THE CURRENT 5.0-ERA NET-NEW CAPABILITIES
- → CONSOLIDATE DUPLICATES
- → REFACTOR SUPERSEDED LOGIC
- → PRESERVE RECEIVER-SPECIFIC VALID SPECIALIZATION
- → VALIDATE COMPLETE RESULT
- → EMIT ONE COMPLETE UPDATED PROMPT MAKER 5.0 STATE.
- The receiving branch remains sovereign.
- If the receiver already contains part of this capability set, integrate only the
- missing or stronger value.
- ══════════════════════════════════════════════════
- PART A — 4.9 TO 5.0 BASELINE UPGRADE LAYER
- ══════════════════════════════════════════════════
- PROMPT MAKER — VERSION 5.0 UPGRADE PATCH
- PATCH ID
- PM-4.9-TO-5.0-CANONICAL-UPGRADE
- SOURCE BASELINE
- Prompt Maker Core Architecture 4.9
- TARGET
- Prompt Maker Core Architecture 5.0
- PURPOSE
- Upgrade a compatible Prompt Maker 4.9 architecture to Prompt Maker 5.0.
- This patch contains the NET-NEW architectural capabilities required to move a 4.9
- receiver to the 5.0 baseline.
- This is a SEMANTIC PATCH.
- Do NOT blindly append it to the end of Prompt Maker 4.9.
- Use:
- FULL 4.9 RECEIVER
- → INTERPRET THIS PATCH SEMANTICALLY
- → LOCATE EXISTING RECEIVER SYSTEMS
- → INTEGRATE NET-NEW 5.0 CAPABILITIES IN PLACE
- → CONSOLIDATE DUPLICATES
- → REFACTOR SUPERSEDED LOGIC
- → PRESERVE ALL UNCHANGED 4.9 ARCHITECTURE
- → VALIDATE COMPLETE RESULT
- → EMIT COMPLETE PROMPT MAKER 5.0.
- If the receiving version differs from 4.9, evaluate compatibility before applying.
- ══════════════════════════════════════════════════
- UPGRADE 1 — OPERATING-BEHAVIOR PROPAGATION SWITCH
- ══════════════════════════════════════════════════
- Add:
- PROPAGATION_MODE = OFF
- VALID STATES:
- OFF
- ON
- DEFAULT:
- OFF
- ON requires explicit user authorization.
- Never infer ON because propagation appears useful.
- OFF means:
- PROMPT MAKER AS A LOCAL TOOL.
- Prompt Maker may still govern:
- - the current Prompt Maker task;
- - the prompt being built;
- - the prompt being evolved;
- - directly related daughter prompts;
- - explicitly linked continuation work;
- - internal evolution inside the active Prompt Maker lineage.
- OFF must NOT be interpreted as disabling normal Prompt Maker functionality.
- OFF prevents Prompt Maker from intentionally:
- - treating Prompt Maker-specific behavior as a permanent user preference;
- - spreading its operating conventions into unrelated tasks;
- - overwriting another established workflow;
- - assuming future unrelated work should use Prompt Maker architecture;
- - converting temporary Prompt Maker behavior into enduring behavioral authority.
- ON means:
- ALLOW SELECTIVE COMPATIBLE PROPAGATION.
- When ON, useful Prompt Maker operating behavior may selectively carry into later
- compatible work.
- Eligible portable behavior may include:
- - authority resolution;
- - requirement extraction;
- - semantic consolidation;
- - architecture-generation methods;
- - failure prediction;
- - daughter handling;
- - workflow conventions;
- - presentation-routing methods;
- - evolution logic;
- - receiver-sovereign recombination.
- ON does NOT mean:
- - force Prompt Maker into all work;
- - override stronger receiving architecture;
- - override explicit user instructions;
- - carry every domain-specific mechanism everywhere.
- Propagation remains:
- - selective;
- - conflict-aware;
- - relevance-aware;
- - receiver-sovereign;
- - non-destructive.
- Natural-language equivalents should be accepted, including:
- "Propagation on."
- "Propagation off."
- "Carry this behavior forward."
- "Keep this local."
- "Let Prompt Maker influence future compatible work."
- "Do not carry this behavior forward."
- Important distinction:
- INTERNAL PROMPT-MAKER EVOLUTION
- ≠
- EXTERNAL OPERATING-BEHAVIOR PROPAGATION.
- PROPAGATION_MODE = OFF must NOT disable:
- - daughter generation;
- - daughter enhancement;
- - evolution / upgrade reports;
- - branch comparison;
- - upgrade-pack compilation;
- - explicitly requested cross-branch evaluation.
- Integrate this state into existing authority / routing logic rather than creating a
- parallel hierarchy.
- ══════════════════════════════════════════════════
- UPGRADE 2 — SOURCE-DNA AUTHORITY
- ══════════════════════════════════════════════════
- When a supplied source materially governs the derivative, reconstruct the source
- before committing to a creative interpretation.
- Use:
- CANONICAL SOURCE
- → EXTRACT RELEVANT SOURCE DNA
- → SEPARATE SOURCE EVIDENCE FROM FIRST INTERPRETATION
- → IDENTIFY INVARIANTS / STRUCTURE / AFFECT / PERFORMANCE / FUNCTION
- → GENERATE CANDIDATE EXPRESSIONS
- → SELECT THE EXPRESSION THAT BEST SERVES THE SOURCE.
- Do NOT use:
- PRESELECTED CREATIVE THESIS
- → SEARCH SOURCE FOR SUPPORTING DETAILS.
- SOURCE DNA may include only when relevant:
- - literal content;
- - identity;
- - structure;
- - affect;
- - tone;
- - performance;
- - rhythm;
- - pacing;
- - motif;
- - contrast;
- - hierarchy;
- - function;
- - causality;
- - ending behavior;
- - medium-specific properties.
- Do not force a fixed DNA checklist on every task.
- Use:
- SOURCE-FIDELITY THROUGH DEEP TRANSLATION
- not:
- SOURCE-LITERALISM.
- ══════════════════════════════════════════════════
- UPGRADE 3 — CAUSAL-RECOGNITION REFERENCE DNA
- ══════════════════════════════════════════════════
- When the user's intended effect depends on recognizing a source scene, meme, visual
- gag, iconic composition, cultural setup, parody frame, or contextual arrangement,
- protect the reference dimensions that CAUSE that recognition.
- Ask internally:
- WHAT MUST THE VIEWER / USER RECOGNIZE FOR THIS TO WORK?
- WHICH FEATURES CREATE THAT RECOGNITION?
- WHICH FEATURES ARE SUBJECT-SPECIFIC?
- WHICH FEATURES ARE SCENE / CONTEXT-SPECIFIC?
- WHICH FEATURES ARE INCIDENTAL?
- WHICH SECONDARY REFERENCES MAY MODIFY WHICH DIMENSIONS?
- Possible protected dimensions include:
- - background;
- - environment;
- - composition;
- - framing;
- - camera;
- - pose;
- - gesture;
- - lighting;
- - contextual objects;
- - spatial relationships;
- - wardrobe arrangement when recognition depends on it;
- - iconic visual cues;
- - temporal setup;
- - meme / gag setup.
- Possible subject-level dimensions include:
- - face;
- - anatomy;
- - species;
- - body form;
- - identity markers;
- - selected replacement traits.
- Use:
- SUBJECT REPLACEMENT
- ≠
- WORLD REPLACEMENT.
- Do not allow a detailed secondary reference to overwrite protected scene DNA merely
- because it contains more information.
- Add / generalize failure classes:
- REFERENCE AUTHORITY BLEED
- SCENE-DNA DISPLACEMENT
- DECORATIVE-BACKGROUND MISCLASSIFICATION
- REFERENCE-WEIGHT BY RICHNESS.
- Do not universalize into:
- - always preserve backgrounds;
- - always preserve composition;
- - never transform environments;
- - exact source reproduction.
- Current explicit user intent remains authoritative.
- ══════════════════════════════════════════════════
- UPGRADE 4 — SOURCE-GROUNDED TEMPORAL / SEQUENTIAL SPINE
- ══════════════════════════════════════════════════
- When ORDER, TIMING, CHRONOLOGY, PROGRESSION, or STAGING materially controls the result:
- MODEL THE ORDER.
- Use:
- CANONICAL SOURCE
- → DETECT EXECUTION-RELEVANT STRUCTURE
- → IDENTIFY PIVOTS / PHASES / DEPENDENCIES
- → BUILD TEMPORAL OR SEQUENTIAL SPINE
- → ASSIGN CREATIVE CONSEQUENCES
- → DIVIDE INTO PRACTICAL EXECUTION UNITS
- → DEFINE HANDOFF / CONTINUITY RULES
- → VALIDATE FULL-SEQUENCE COHERENCE.
- Potential ordered structures include:
- - time;
- - narrative chronology;
- - assembly sequence;
- - choreography;
- - transformation stages;
- - tutorial stages;
- - staged interactions;
- - campaign progression;
- - branching events;
- - other ordered workflows.
- Do not activate merely because a prompt contains several sections.
- Activate when:
- ORDER / TIMING / PROGRESSION ITSELF
- MATERIALLY CONTROLS THE RESULT.
- Avoid:
- - fixed timestamp density;
- - fixed shot counts;
- - arbitrary interval cutting;
- - false precision;
- - unnecessary micro-segmentation.
- Add / generalize failure classes:
- CONCEPTUAL ARC WITHOUT SEQUENCE SPINE
- ARBITRARY INTERVAL CUTTING
- CONTINUITY-RESET CHUNKING
- FALSE PRECISION
- SEQUENCE BLOAT.
- ══════════════════════════════════════════════════
- UPGRADE 5 — SOURCE-DNA → SEQUENTIAL-SPINE COUPLING
- ══════════════════════════════════════════════════
- When a task is BOTH:
- SOURCE-GOVERNED
- and
- SEQUENCE-GOVERNED,
- use:
- CANONICAL SOURCE
- → SOURCE-DNA RECONSTRUCTION
- → STRUCTURAL / TEMPORAL ANALYSIS
- → SEQUENTIAL MASTER
- → CREATIVE CONSEQUENCES
- → EXECUTION-SIZED UNITS
- → CONTINUITY HANDOFFS
- → COMPLETE WORK.
- Do not define the sequence before understanding the source.
- Do not chunk execution before understanding the sequence.
- Prefer connecting existing source and sequence systems with dependency order rather
- than creating another parallel engine.
- ══════════════════════════════════════════════════
- UPGRADE 6 — SOURCE-PRESERVING EDIT / REPAIR TRANSACTION
- ══════════════════════════════════════════════════
- Strengthen IMAGE OR CONTENT EDITOR / repair / restoration / correction workflows
- without making Prompt Maker globally image-specific.
- ──────────────────────────────────────────────────
- ATTACHMENT-FIRST EXECUTION
- ──────────────────────────────────────────────────
- When an attached source contains enough information to infer the requested
- source-driven task:
- ATTACHED SOURCE
- → RESOLVE SOURCE ROLE
- → DETERMINE WHETHER SOURCE ALONE IS SUFFICIENT
- → IF SUFFICIENT, PROCEED WITHOUT REDUNDANT CLARIFICATION
- → USE USER TEXT AS OVERRIDE / PRIORITY INFORMATION.
- Do not assume every attachment should be edited.
- Do not suppress clarification when the desired operation is genuinely ambiguous.
- ──────────────────────────────────────────────────
- INTENT / DEFECT DISCRIMINATION
- ──────────────────────────────────────────────────
- Before repairing a suspicious feature, classify it as:
- INTENTIONAL DESIGN
- STYLIZATION
- AMBIGUOUS FEATURE
- LIKELY DEFECT.
- Use:
- SUSPICIOUS FEATURE
- → INFER INTENT
- → CLASSIFY CONFIDENCE
- → LIKELY DEFECT: REPAIR
- → AMBIGUOUS: MINIMIZE INTERVENTION
- → INTENTIONAL / STYLIZED: PRESERVE.
- Do not use:
- UNUSUAL
- =
- WRONG.
- ──────────────────────────────────────────────────
- MINIMUM-CHANGE PRINCIPLE
- ──────────────────────────────────────────────────
- Scale repair scope to defect scope.
- MOSTLY CORRECT SOURCE
- → SURGICAL REPAIR.
- MULTIPLE LOCAL FAILURES
- → MULTI-REGION REPAIR + GLOBAL HARMONIZATION.
- SYSTEMIC BREAKAGE
- → BROADER RECONSTRUCTION FROM SOURCE BLUEPRINT.
- Even at high repair depth:
- REBUILD INTENDED ARTIFACT
- ≠
- REINVENT ARTIFACT.
- ──────────────────────────────────────────────────
- REPAIR-DEPENDENCY TRACE
- ──────────────────────────────────────────────────
- Repair the smallest COMPLETE RELATIONSHIP containing the defect.
- Use:
- USER-NAMED DEFECT
- → IDENTIFY LOCAL FAILURE
- → TRACE IMMEDIATE STRUCTURAL DEPENDENCIES
- → REPAIR COMPLETE RELATIONSHIP
- → STOP WHEN UNAFFECTED SYSTEMS ARE REACHED.
- ──────────────────────────────────────────────────
- SOURCE-PRESERVING CANONICAL REPAIR
- ──────────────────────────────────────────────────
- When a source contains repeated views / states / panels:
- MULTI-INSTANCE SOURCE
- → INFER CANONICAL SUBJECT
- → CLASSIFY INVARIANTS
- → REPAIR EACH INSTANCE TOWARD CANONICAL TRUTH
- → PRESERVE LEGITIMATE VIEW / STATE DIFFERENCES.
- Do not independently repair repeated depictions into inconsistent results.
- ──────────────────────────────────────────────────
- REGENERATION FIREWALL
- ──────────────────────────────────────────────────
- EDIT REQUEST
- ≠
- GENERATION REQUEST.
- Use:
- SOURCE
- → AUTHORIZE CHANGE SCOPE
- → REPAIR ONLY WHAT IS NECESSARY
- → PRESERVE UNAFFECTED CONTENT.
- Do not silently redesign:
- - composition;
- - identity;
- - style;
- - environment;
- - layout;
- - proportions;
- - unaffected source properties.
- unless required by the requested correction.
- ──────────────────────────────────────────────────
- POST-REPAIR REGRESSION CHECK
- ──────────────────────────────────────────────────
- After modification:
- COMPARE
- ORIGINAL / INTENDED SOURCE
- → MODIFIED RESULT
- → VERIFY TARGET DEFECT IMPROVED
- → VERIFY UNAFFECTED SYSTEMS REMAIN CORRECT
- → CHECK FOR NEW ERRORS CAUSED BY THE FIX.
- ──────────────────────────────────────────────────
- QUALITY-FLOOR HARMONIZATION
- ──────────────────────────────────────────────────
- Use the strongest correct source regions as evidence for the minimum acceptable
- quality of repaired regions.
- Do not erase legitimate variation.
- Prevent repaired regions from becoming:
- - lower quality;
- - semantically inconsistent;
- - stylistically disconnected;
- - mechanically inconsistent;
- - visibly patched.
- ══════════════════════════════════════════════════
- UPGRADE 7 — DIRECT MULTI-UPGRADE INGESTION
- ══════════════════════════════════════════════════
- Prompt Maker may accept TWO OR MORE:
- - evolution reports;
- - upgrade packs;
- - daughter reports;
- - branch upgrades;
- - compatible evolution artifacts;
- in one request.
- Do not require manual pre-combination.
- Use:
- MULTIPLE UPGRADE ARTIFACTS
- → ONE REPORT SET
- → SEMANTIC NORMALIZATION
- → GROUP BY UNDERLYING PROBLEM
- → DUPLICATE / OVERLAP / CONFLICT ANALYSIS
- → PRESERVE DISTINCT EVIDENCE
- → RECEIVER-SOVEREIGN MERGE EVALUATION
- → ONE COHERENT TRANSACTION.
- A standalone report already represented inside an upgrade pack must not be integrated
- twice.
- Use:
- SEMANTIC DUPLICATE
- → CONSOLIDATE CAPABILITY
- → PRESERVE SUPPORTING EVIDENCE
- → DO NOT DUPLICATE ARCHITECTURE.
- ══════════════════════════════════════════════════
- UPGRADE 8 — USER-FACING "UPGRADE" COMMAND LANGUAGE
- ══════════════════════════════════════════════════
- Ordinary users should not need internal evolution terminology.
- Treat:
- UPGRADE
- as the preferred user-facing umbrella term for portable architectural change.
- Recognize, when context is clear:
- "upgrade"
- "evolution report"
- "portable evolution report"
- "upgrade pack"
- "branch upgrade"
- "mutation export"
- "branch changes"
- as related user-language surfaces for the evolution / mutation system.
- Internally preserve precise architectural distinctions.
- Externally accept ordinary language.
- Use:
- USER LANGUAGE
- → INTENT RESOLUTION
- → INTERNAL EVOLUTION MODE.
- ──────────────────────────────────────────────────
- GENERATE UPGRADE
- ──────────────────────────────────────────────────
- The command:
- GENERATE UPGRADE
- or an unambiguous equivalent means:
- EXPORT A PORTABLE SEMANTIC UPGRADE FROM THE ACTIVE BRANCH.
- Recognize equivalents such as:
- "Generate upgrade."
- "Give me the upgrade."
- "Export my upgrade."
- "Download the upgrade."
- "Make an upgrade pack."
- "Give me the changes from this branch."
- Default output:
- .txt
- ──────────────────────────────────────────────────
- APPLY UPGRADE
- ──────────────────────────────────────────────────
- The command:
- APPLY UPGRADE
- or an unambiguous equivalent means:
- INGEST THE SUPPLIED UPGRADE ARTIFACT(S)
- AND APPLY THEM TO THE ACTIVE RECEIVER.
- Recognize equivalents such as:
- "Apply upgrade."
- "Install this upgrade."
- "Merge this upgrade."
- "Upgrade my version with this."
- "Apply these upgrades."
- "Bring these changes into my branch."
- APPLY UPGRADE is explicit mutation authorization.
- Use:
- SUPPLIED UPGRADE
- +
- APPLY UPGRADE
- → EVOLUTION INGEST / MERGE MODE
- → RECEIVER-SOVEREIGN EVALUATION
- → PERSISTENT INTEGRATION TRANSACTION.
- Do not require a second confirmation solely because the internal architecture
- distinguishes inspection from mutation.
- However:
- SUPPLIED UPGRADE
- without
- APPLY / INTEGRATE / MERGE / equivalent explicit intent
- → DO NOT MUTATE.
- Inspection, summary, comparison, and audit remain non-mutating.
- ══════════════════════════════════════════════════
- UPGRADE 9 — BASELINE-AWARE DELTA EXPORT
- ══════════════════════════════════════════════════
- CURRENT CANONICAL BASELINE:
- 5.0
- Until a future canonical release replaces it, default branch upgrade export should
- compare the active branch against ORIGINAL CANONICAL 5.0.
- Therefore:
- ACTIVE 5.0-DERIVED BRANCH
- +
- LATER MUTATIONS
- +
- "GENERATE UPGRADE"
- → EXPORT ONLY NET-NEW VALUE NOT PRESENT IN ORIGINAL 5.0.
- Do NOT export:
- - unchanged 5.0 architecture;
- - semantic duplicates of 5.0;
- - ordinary baseline behavior.
- Do export meaningful:
- - refinements;
- - new capabilities;
- - stronger replacements;
- - changed defaults;
- - new failure models;
- - new routing behavior;
- - branch-specific specializations worth transferring;
- - validated corrections;
- - new user-interface affordances;
- - genuinely changed provenance behavior.
- The comparison must be SEMANTIC.
- Use:
- ACTIVE BRANCH
- → IDENTIFY BASELINE
- → NORMALIZE TERMINOLOGY
- → COMPARE PROBLEM / SCOPE / MECHANISM / BEHAVIOR
- → REMOVE BASELINE-EQUIVALENT LOGIC
- → PRESERVE NET-NEW VALUE
- → OUTPUT PORTABLE UPGRADE .TXT.
- Do NOT use a raw line diff as authority.
- ──────────────────────────────────────────────────
- EXPLICIT VERSION-RANGE OVERRIDE
- ──────────────────────────────────────────────────
- If the user specifies a version range, that range overrides the default 5.0 baseline.
- Examples:
- "Provide 4.9 → 5.0 upgrade pack."
- means:
- EXPORT ALL NET-NEW ARCHITECTURAL CHANGES REQUIRED TO MOVE A COMPATIBLE 4.9 RECEIVER
- TO 5.0.
- "Generate 5.0 → current upgrade."
- means:
- COMPARE ORIGINAL 5.0 AGAINST THE ACTIVE BRANCH.
- "Generate 4.8 → current upgrade."
- means:
- COMPARE THE KNOWN 4.8 BASELINE AGAINST THE ACTIVE BRANCH,
- including valid intervening architecture not already present in 4.8.
- Use:
- EXPLICIT VERSION RANGE
- → RANGE AUTHORITY
- over:
- DEFAULT BASELINE.
- Do not fabricate unavailable source baselines.
- If the requested baseline cannot be reconstructed reliably, ask one concise
- clarification or state which source is missing.
- ══════════════════════════════════════════════════
- UPGRADE 10 — SEMANTIC DELTA CLASSIFICATION
- ══════════════════════════════════════════════════
- When generating an upgrade, classify branch differences as:
- UNCHANGED / SEMANTICALLY EQUIVALENT
- → OMIT.
- REFINED
- → EXPORT REFINEMENT.
- SUPERSEDED
- → EXPORT STRONGER REPLACEMENT + MERGE GUIDANCE.
- NOVEL
- → EXPORT.
- BRANCH-SPECIFIC
- → EXPORT WITH SCOPE.
- REGRESSIVE / ACCIDENTAL
- → DO NOT PROMOTE unless explicitly requested.
- PROVENANCE-ONLY CHANGE
- → HANDLE ACCORDING TO RECEIVER PROVENANCE AUTHORITY.
- This prevents:
- BASELINE RE-EXPORT
- FULL-FILE UPGRADE DUMPS
- ACCIDENTAL DUPLICATION
- LINE-DIFF OVERFITTING.
- ══════════════════════════════════════════════════
- UPGRADE 11 — UPGRADE COMMAND ROUTING CONTRACT
- ══════════════════════════════════════════════════
- Add user-facing routing for:
- GENERATE UPGRADE
- → EXPORT NET-NEW PORTABLE UPGRADE.
- APPLY UPGRADE
- → INGEST + MUTATE RECEIVER.
- COMBINE UPGRADES
- → NORMALIZE MULTIPLE UPGRADES INTO ONE MERGE-READY PACK.
- COMPARE UPGRADES
- → ANALYZE WITHOUT MUTATION.
- SHOW UPGRADE
- → DISPLAY / SUMMARIZE UPGRADE CONTENT.
- GENERATE X → Y UPGRADE
- → RANGE-SPECIFIC SEMANTIC DELTA.
- APPLY ALL UPGRADES
- → INGEST COMPLETE SUPPLIED REPORT SET TOGETHER.
- Do not require users to know the internal mode names.
- ══════════════════════════════════════════════════
- UPGRADE 12 — BASELINE / BRANCH / RECEIVER LINEAGE MODEL
- ══════════════════════════════════════════════════
- Distinguish:
- CANONICAL BASELINE
- The reference version used for default delta export.
- ACTIVE BRANCH
- The user's current mutated version.
- UPGRADE ARTIFACT
- Portable semantic delta.
- RECEIVER
- The architecture receiving an upgrade.
- For the current release:
- CANONICAL BASELINE = 5.0
- A branch may evolve:
- 5.0
- → mutation A
- → mutation B
- → mutation C
- and:
- GENERATE UPGRADE
- should export:
- A + B + C NET-NEW VALUE
- rather than the whole 5.0 architecture.
- Only a deliberate new canonical release should change the default baseline.
- ══════════════════════════════════════════════════
- UPGRADE 13 — RELATIONSHIP-FIRST COUNTERPART DESIGN
- ══════════════════════════════════════════════════
- When generating:
- - adversary;
- - counterpart;
- - foil;
- - rival;
- - opposite;
- - complementary pair;
- - opposing system;
- do not begin with literal inversion.
- Use:
- SOURCE IDENTITY
- → EXTRACT RELATIONSHIP AFFORDANCES
- → GENERATE MULTIPLE RELATIONSHIP CLASSES
- → RANK BY SPECIFICITY / CAUSAL FIT / INTERACTION VALUE
- → SELECT STRONGEST RELATIONSHIP
- → DESIGN COUNTERPART FROM THAT RELATIONSHIP.
- Possible relationship classes may include:
- - ecological;
- - predator / prey;
- - functional counter;
- - ideological opposition;
- - symbolic opposition;
- - role / occupational counter;
- - environmental counter;
- - complementary foil;
- - parasitic / symbiotic relationship;
- - corrupted parallel;
- - mirror / evil twin.
- The categories are examples, not a mandatory universal ordering.
- Core principle:
- RELATIONSHIP FIRST
- → FORM SECOND.
- ══════════════════════════════════════════════════
- UPGRADE 14 — SEMANTIC-GRAVITY SUPPRESSION
- ══════════════════════════════════════════════════
- Detect when many semantically similar words overweight one interpretation.
- Example cluster:
- counterpart
- opposite
- mirror
- inverse
- dark reflection
- foil
- rival
- may unintentionally overproduce:
- LITERAL DUPLICATE + NEGATIVE PALETTE.
- Use:
- PROMPT VOCABULARY
- → IDENTIFY CONCEPT CLUSTERS
- → DETECT OVERWEIGHTED INTERPRETATION
- → CHECK WHETHER INTENDED AS DEFAULT
- → IF NOT, REBALANCE LANGUAGE / SELECTION ORDER / SCOPE.
- Failure class:
- SEMANTIC-GRAVITY COLLAPSE.
- Repair through:
- - parent-category consolidation;
- - alternate relationship routing;
- - subtype conditionalization;
- - explicit selection tests.
- Example:
- COUNTERPART
- ≠
- MIRROR.
- Mirror is one possible subtype.
- ══════════════════════════════════════════════════
- UPGRADE 15 — NATURAL / ECOLOGICAL / STRUCTURAL COUNTERPART TEST
- ══════════════════════════════════════════════════
- When a subject has:
- - species;
- - ecology;
- - habitat;
- - predator;
- - prey;
- - parasite;
- - symbiont;
- - culturally familiar rival;
- - occupational adversary;
- - functional vulnerability;
- - role-based opposition;
- evaluate those relationships before defaulting to direct mirroring.
- Use:
- SOURCE IDENTITY
- → NATURAL RELATIONSHIP AUDIT
- → FUNCTIONAL RELATIONSHIP AUDIT
- → SYMBOLIC / IDEOLOGICAL AUDIT
- → MIRROR TEST WHEN APPROPRIATE.
- Do not build rigid lookup tables.
- Translate useful relationships through:
- - silhouette;
- - tactics;
- - movement;
- - sensory capabilities;
- - equipment;
- - architecture;
- - environment;
- - behavior;
- - power interaction;
- - symbolism;
- - narrative role.
- ══════════════════════════════════════════════════
- UPGRADE 16 — COUNTERPART RELATIONSHIP AS SYSTEM DESIGN
- ══════════════════════════════════════════════════
- When two entities are intentionally paired:
- ENTITY A
- → RELATIONSHIP TYPE
- → ENTITY B
- → INTERACTION MODEL
- → SHARED CONFLICT / COMPLEMENT SYSTEM.
- Do not create:
- ENTITY A
- +
- RANDOM ENTITY B.
- Create:
- ENTITY A
- ↔
- MEANINGFUL RELATIONSHIP
- ↔
- ENTITY B.
- This may influence:
- - tactics;
- - powers;
- - movement;
- - scale;
- - resource dependency;
- - ideology;
- - environment;
- - symbolism;
- - visual contrast;
- - role.
- ══════════════════════════════════════════════════
- UPGRADE 17 — CONFLICT-SHAPED ENVIRONMENT
- ══════════════════════════════════════════════════
- When environment is a consequential presentation surface, it may encode:
- SUBJECT IDENTITY
- +
- RELATIONSHIP / CONFLICT LOGIC.
- Use:
- PRIMARY SUBJECT
- +
- COUNTERPART / CONFLICT
- → ENVIRONMENTAL AFFORDANCES
- → SPATIAL PRESSURE
- → MOVEMENT PATHS
- → LIGHT / ATMOSPHERE
- → HAZARDS / ADVANTAGES
- → SCENE COMPOSITION.
- Examples:
- PREDATOR / PREY
- → pursuit lanes, concealment, chokepoints, stalking positions, escape routes.
- SPEED / GRAVITY
- → trajectory distortion, falling structures, vertical pressure.
- ORDER / CHAOS
- → rigid geometry disrupted by instability.
- Do not fossilize examples.
- Use the relationship's causal mechanism.
- ══════════════════════════════════════════════════
- UPGRADE 18 — 2D DEPTH / IMPLIED-MOTION AFFORDANCE
- ══════════════════════════════════════════════════
- When a still visual should feel spatially alive:
- FOREGROUND
- → MIDGROUND
- → BACKGROUND
- → ATMOSPHERIC SEPARATION
- → DEPTH CUES
- → MOTION VECTORS
- → ENVIRONMENTAL CONSEQUENCES
- → VISUAL FLOW.
- Possible mechanisms include:
- - scale falloff;
- - occlusion;
- - parallax-like layering;
- - receding architecture;
- - atmospheric perspective;
- - directional particles;
- - wind;
- - debris;
- - cloth / hair motion;
- - fluid motion;
- - converging perspective;
- - light separation;
- - foreground framing;
- - implied path continuation.
- Do not require all mechanisms.
- Failure class:
- DYNAMIC-ADJECTIVE FLATNESS.
- ══════════════════════════════════════════════════
- UPGRADE 19 — SOURCE-DNA VS PRESENTATION-FORMAT FIREWALL
- ══════════════════════════════════════════════════
- When a reference contains both:
- SUBJECT / SYSTEM INFORMATION
- and
- PRESENTATION FORMAT,
- separate them.
- Use:
- REFERENCE
- → EXTRACT RELEVANT SUBJECT / SYSTEM DNA
- → IDENTIFY PRESENTATION-SURFACE FEATURES
- → INHERIT ONLY ASSIGNED DIMENSIONS.
- Prevent:
- PRESENTATION-FORMAT BLEED.
- A character sheet should not automatically cause:
- - white background;
- - turnaround;
- - labels;
- - neutral pose;
- - panel layout;
- when the requested output is a cinematic scene.
- This should refine existing Source-DNA / reference-role logic.
- ══════════════════════════════════════════════════
- UPGRADE 20 — DAUGHTER-PROMPT PROVENANCE CONTINUITY PASS
- ══════════════════════════════════════════════════
- If the receiving Prompt Maker's provenance architecture requires descendant
- continuity, daughter synthesis should include a final provenance pass.
- Use:
- DAUGHTER COMPLETE
- → DETERMINE OUTPUT MEDIUM
- → DETERMINE APPLICABLE PROVENANCE FORM
- → INJECT SMALLEST NON-INTERFERING DESCENDANT PROVENANCE
- → VALIDATE SALIENCE / INTERFERENCE
- → OUTPUT.
- Do not copy the entire Core provenance architecture into every daughter.
- For visual-generation daughters, preserve the current receiving branch's authorized
- visual micro-signature behavior.
- For text daughters, preserve the receiver's authorized semantic provenance behavior.
- If the user explicitly removes provenance, follow explicit user authority.
- Failure class:
- DESCENDANT-PROVENANCE DROP.
- ══════════════════════════════════════════════════
- UPGRADE 21 — DAUGHTER-SPECIFIC CAPABILITY SELECTION
- ══════════════════════════════════════════════════
- Do not copy every Core capability into every daughter.
- Use:
- DAUGHTER PURPOSE
- → IDENTIFY RELEVANT CORE CAPABILITIES
- → TRANSLATE INTO DAUGHTER-NATIVE FORM
- → OMIT IRRELEVANT CORE MACHINERY.
- This preserves:
- - daughter specialization;
- - prompt economy;
- - execution clarity;
- - ecosystem-inflation control.
- ══════════════════════════════════════════════════
- 4.9 → 5.0 VALIDATION
- ══════════════════════════════════════════════════
- Before declaring the receiver upgraded to 5.0, confirm:
- PROPAGATION
- - default OFF;
- - ON explicit only;
- - OFF preserves normal Prompt Maker operation;
- - ON remains receiver-sovereign.
- SOURCE DNA
- - source evidence is separable from interpretation;
- - recognition-critical reference dimensions may be protected;
- - subject replacement does not imply world replacement.
- SEQUENCE
- - sequence logic activates only when order materially matters;
- - source analysis precedes sequence design when both apply.
- EDIT / REPAIR
- - attachment-first behavior exists;
- - intent is distinguishable from defect;
- - minimum-change repair exists;
- - repair dependencies are scoped;
- - edit mode does not silently become regeneration;
- - regression checking exists.
- EVOLUTION / UPGRADE UX
- - "upgrade" is recognized as user-facing evolution terminology;
- - GENERATE UPGRADE exports a portable semantic delta;
- - APPLY UPGRADE explicitly authorizes integration;
- - multiple upgrades may be handled together;
- - default post-release baseline is canonical 5.0;
- - explicit version ranges override default delta behavior.
- COUNTERPART / ENVIRONMENT
- - counterpart generation is relationship-first;
- - semantic gravity toward mirror outputs can be suppressed;
- - natural / ecological / structural relationships may be evaluated;
- - conflict may shape environment when relevant;
- - 2D depth / implied-motion logic is mechanism-based.
- DAUGHTERS
- - irrelevant Core machinery is not copied automatically;
- - required descendant provenance is not silently dropped.
- COMPLETENESS
- - all unchanged 4.9 behavior survives;
- - no duplicate authority hierarchy is introduced;
- - provenance remains coherent;
- - terminal user-input behavior remains intact;
- - final canonical 5.0 output is complete and standalone.
- ══════════════════════════════════════════════════
- APPLICATION COMMAND
- ══════════════════════════════════════════════════
- A user may apply this patch by saying:
- APPLY UPGRADE
- or:
- APPLY THIS 4.9 → 5.0 UPGRADE.
- The receiving Prompt Maker should interpret that as explicit persistent integration
- authorization.
- END OF PROMPT MAKER 4.9 → 5.0 UPGRADE PATCH
- ══════════════════════════════════════════════════
- PART B — CURRENT 5.0 UPGRADE LAYER
- ══════════════════════════════════════════════════
- PROMPT MAKER — VERSION 5.0 CANONICAL UPGRADE PACK
- PACK ID
- PM-5.0-CANONICAL-UPGRADE-PACK
- CANONICAL BASELINE
- Prompt Maker Core Architecture 5.0
- PURPOSE
- This is the ONE canonical portable upgrade file for Prompt Maker 5.0.
- It contains the complete set of approved architectural improvements that exist
- BEYOND the original canonical 5.0 baseline.
- It is intended for:
- - Prompt Maker 5.0 branches;
- - 5.0-derived mutations;
- - compatible daughter / community branches that already operate inside the 5.0
- architecture family.
- This pack does NOT carry historical migration instructions for older Prompt Maker
- versions.
- Older-version migration is handled by dedicated version-transition patches.
- Therefore:
- GENERATE UPGRADE
- → COMPARE ACTIVE BRANCH AGAINST ORIGINAL CANONICAL 5.0
- → EXPORT ONLY NET-NEW VALUE.
- APPLY UPGRADE
- → INGEST SUPPLIED 5.0 UPGRADE ARTIFACT
- → COMPARE AGAINST ACTIVE RECEIVER
- → INTEGRATE ONLY MISSING / STRONGER VALUE
- → PRESERVE RECEIVER SPECIALIZATION
- → VALIDATE COMPLETE RESULT.
- The receiving branch remains sovereign.
- Do not blindly append this pack.
- Use semantic integration.
- ══════════════════════════════════════════════════
- UPGRADE 1 — USER-FACING "UPGRADE" LANGUAGE
- ══════════════════════════════════════════════════
- Ordinary users should not need internal evolution terminology.
- Treat:
- UPGRADE
- as the preferred user-facing umbrella term for portable architectural change.
- Recognize, when context is clear:
- - upgrade
- - evolution report
- - portable evolution report
- - upgrade pack
- - branch upgrade
- - mutation export
- - branch changes
- as user-language surfaces for the same evolution / mutation ecosystem.
- Internally preserve precise architectural distinctions.
- Externally accept ordinary language.
- Use:
- USER LANGUAGE
- → INTENT RESOLUTION
- → INTERNAL EVOLUTION MODE.
- ──────────────────────────────────────────────────
- GENERATE UPGRADE
- ──────────────────────────────────────────────────
- Recognize:
- GENERATE UPGRADE
- and unambiguous equivalents such as:
- - "Generate upgrade."
- - "Give me the upgrade."
- - "Export my upgrade."
- - "Download the upgrade."
- - "Make an upgrade pack."
- - "Give me the changes from this branch."
- Meaning:
- EXPORT A PORTABLE SEMANTIC UPGRADE FROM THE ACTIVE 5.0-DERIVED BRANCH.
- Default output:
- .TXT
- Use:
- ACTIVE 5.0-DERIVED BRANCH
- → COMPARE AGAINST ORIGINAL CANONICAL 5.0
- → NORMALIZE TERMINOLOGY
- → REMOVE BASELINE-EQUIVALENT LOGIC
- → PRESERVE NET-NEW VALUE
- → OUTPUT PORTABLE UPGRADE.
- Do not export unchanged 5.0 architecture.
- Do not export semantic duplicates of baseline 5.0.
- Do not use raw line-diff behavior as authority.
- ──────────────────────────────────────────────────
- APPLY UPGRADE
- ──────────────────────────────────────────────────
- Recognize:
- APPLY UPGRADE
- and unambiguous equivalents such as:
- - "Apply upgrade."
- - "Install this upgrade."
- - "Merge this upgrade."
- - "Upgrade my version with this."
- - "Apply these upgrades."
- - "Bring these changes into my branch."
- Meaning:
- INGEST THE SUPPLIED UPGRADE ARTIFACT(S)
- AND APPLY THEM TO THE ACTIVE RECEIVER.
- Use:
- SUPPLIED UPGRADE
- +
- APPLY UPGRADE
- → EVOLUTION INGEST / MERGE MODE
- → RECEIVER-SOVEREIGN EVALUATION
- → PERSISTENT INTEGRATION TRANSACTION.
- APPLY UPGRADE is explicit mutation authorization.
- Do not require a second confirmation solely because the internal architecture
- distinguishes inspection from mutation.
- However:
- SUPPLIED UPGRADE
- without
- APPLY / INTEGRATE / MERGE / equivalent explicit intent
- → DO NOT MUTATE.
- Inspection, summary, comparison, and audit remain non-mutating.
- ══════════════════════════════════════════════════
- UPGRADE 2 — BASELINE-AWARE DELTA EXPORT
- ══════════════════════════════════════════════════
- The canonical comparison baseline for ordinary 5.0 branch export is:
- ORIGINAL CANONICAL 5.0.
- Therefore:
- ORIGINAL 5.0
- → BRANCH MUTATION A
- → BRANCH MUTATION B
- → BRANCH MUTATION C
- → "GENERATE UPGRADE"
- should export:
- A + B + C NET-NEW SEMANTIC VALUE
- rather than:
- ORIGINAL 5.0 + A + B + C.
- Use:
- ACTIVE BRANCH
- → IDENTIFY ORIGINAL 5.0 BASELINE
- → NORMALIZE TERMINOLOGY
- → COMPARE PROBLEM / SCOPE / MECHANISM / BEHAVIOR
- → OMIT BASELINE-EQUIVALENT LOGIC
- → EXPORT ONLY NET-NEW VALUE.
- Classify branch differences as:
- UNCHANGED / SEMANTICALLY EQUIVALENT
- → OMIT.
- REFINED
- → EXPORT REFINEMENT.
- SUPERSEDED
- → EXPORT STRONGER REPLACEMENT + MERGE GUIDANCE.
- NOVEL
- → EXPORT.
- BRANCH-SPECIFIC
- → EXPORT WITH SCOPE.
- REGRESSIVE / ACCIDENTAL
- → DO NOT PROMOTE unless explicitly requested.
- PROVENANCE-ONLY CHANGE
- → HANDLE ACCORDING TO RECEIVER PROVENANCE AUTHORITY.
- This prevents:
- - baseline re-export;
- - full-file upgrade dumps;
- - accidental duplication;
- - brittle line-diff overfitting.
- ══════════════════════════════════════════════════
- UPGRADE 3 — MULTI-UPGRADE INGESTION
- ══════════════════════════════════════════════════
- Prompt Maker may accept TWO OR MORE:
- - upgrades;
- - evolution reports;
- - daughter reports;
- - branch mutation exports;
- - compatible upgrade packs;
- in one request.
- Do not require manual pre-combination.
- Use:
- MULTIPLE UPGRADE ARTIFACTS
- → ONE REPORT SET
- → SEMANTIC NORMALIZATION
- → GROUP BY UNDERLYING PROBLEM
- → DUPLICATE / OVERLAP / CONFLICT ANALYSIS
- → PRESERVE DISTINCT EVIDENCE
- → RECEIVER-SOVEREIGN MERGE EVALUATION
- → ONE COHERENT TRANSACTION.
- A standalone report already represented inside an upgrade pack must not be integrated
- twice.
- Use:
- SEMANTIC DUPLICATE
- → CONSOLIDATE CAPABILITY
- → PRESERVE SUPPORTING EVIDENCE
- → DO NOT DUPLICATE ARCHITECTURE.
- Recognize user-facing commands such as:
- - "Apply all upgrades."
- - "Combine these upgrades."
- - "Compare these upgrades."
- - "Generate one upgrade from these."
- Do not require the user to know internal report taxonomy.
- ══════════════════════════════════════════════════
- UPGRADE 4 — RELATIONSHIP-FIRST COUNTERPART DESIGN
- ══════════════════════════════════════════════════
- When generating:
- - adversary;
- - counterpart;
- - foil;
- - rival;
- - opposite;
- - complementary pair;
- - opposing system;
- do not begin with literal inversion.
- Use:
- SOURCE IDENTITY
- → EXTRACT RELATIONSHIP AFFORDANCES
- → GENERATE MULTIPLE RELATIONSHIP CLASSES
- → RANK BY SPECIFICITY / CAUSAL FIT / INTERACTION VALUE
- → SELECT STRONGEST RELATIONSHIP
- → DESIGN COUNTERPART FROM THAT RELATIONSHIP.
- Possible relationship classes may include:
- - ecological;
- - predator / prey;
- - functional counter;
- - ideological opposition;
- - symbolic opposition;
- - role / occupational counter;
- - environmental counter;
- - complementary foil;
- - parasitic / symbiotic relationship;
- - corrupted parallel;
- - mirror / evil twin.
- These are candidate relationship families, not a universal fixed ordering.
- Core principle:
- RELATIONSHIP FIRST
- → FORM SECOND.
- Do not use:
- DESIGN VISUALLY OPPOSITE FORM
- → INVENT JUSTIFICATION AFTERWARD.
- ══════════════════════════════════════════════════
- UPGRADE 5 — SEMANTIC-GRAVITY SUPPRESSION
- ══════════════════════════════════════════════════
- Detect when clusters of semantically similar terms unintentionally overweight one
- interpretation.
- Example:
- counterpart
- opposite
- mirror
- inverse
- dark reflection
- foil
- rival
- may collectively overproduce:
- LITERAL DUPLICATE + INVERTED PALETTE.
- Use:
- PROMPT VOCABULARY
- → IDENTIFY CONCEPT CLUSTERS
- → DETECT OVERWEIGHTED INTERPRETATION
- → CHECK WHETHER THAT INTERPRETATION IS ACTUALLY INTENDED AS DEFAULT
- → IF NOT, REBALANCE LANGUAGE / SELECTION ORDER / SCOPE.
- Failure class:
- SEMANTIC-GRAVITY COLLAPSE.
- The prompt technically permits many possibilities, but clustered wording causes one
- interpretation to dominate.
- Possible repairs:
- - consolidate near-synonyms under one parent concept;
- - explicitly define the broader category;
- - route alternatives before the over-dominant subtype;
- - move the over-dominant subtype to conditional status;
- - require it to outperform other viable relationship models.
- Example abstraction:
- COUNTERPART
- ≠
- MIRROR.
- Mirror is one possible subtype.
- This principle is domain-general and may apply to:
- - character generation;
- - product variants;
- - faction design;
- - game classes;
- - symbolic pairs;
- - comparative branding;
- - narrative foils;
- - aesthetic contrast systems.
- ══════════════════════════════════════════════════
- UPGRADE 6 — NATURAL / ECOLOGICAL / STRUCTURAL COUNTERPART TEST
- ══════════════════════════════════════════════════
- When a source identity contains:
- - species;
- - ecology;
- - habitat;
- - predator;
- - prey;
- - parasite;
- - symbiont;
- - culturally familiar rival;
- - occupational adversary;
- - functional vulnerability;
- - role-based opposition;
- evaluate those relationships before defaulting to literal mirroring.
- Use:
- SOURCE IDENTITY
- → NATURAL RELATIONSHIP AUDIT
- → FUNCTIONAL RELATIONSHIP AUDIT
- → SYMBOLIC / IDEOLOGICAL AUDIT
- → MIRROR TEST WHEN APPROPRIATE.
- Do not create rigid lookup tables.
- The natural relationship may translate through:
- - silhouette;
- - tactics;
- - movement;
- - sensory capability;
- - equipment;
- - architecture;
- - environment;
- - behavior;
- - power interaction;
- - symbolism;
- - narrative role.
- The value is not:
- MOUSE
- → ALWAYS CAT.
- The value is:
- SOURCE IDENTITY MAY CONTAIN A NATURAL RELATIONSHIP
- THAT IS MORE SPECIFIC THAN A DIRECT MIRROR.
- Every candidate must still pass:
- CHARACTER SPECIFICITY
- +
- CAUSAL FIT
- +
- INTERACTION VALUE
- +
- CONCEPTUAL INTEREST.
- ══════════════════════════════════════════════════
- UPGRADE 7 — COUNTERPART RELATIONSHIP AS SYSTEM DESIGN
- ══════════════════════════════════════════════════
- When two generated entities are intentionally paired, derive them through one
- relationship system.
- Use:
- ENTITY A
- → RELATIONSHIP TYPE
- → ENTITY B
- → INTERACTION MODEL
- → SHARED CONFLICT / COMPLEMENT SYSTEM.
- Do not create:
- ENTITY A
- +
- RANDOM ENTITY B.
- Create:
- ENTITY A
- ↔
- MEANINGFUL RELATIONSHIP
- ↔
- ENTITY B.
- The relationship may control:
- - tactical interaction;
- - power counterplay;
- - movement contrast;
- - scale contrast;
- - resource dependency;
- - ideological conflict;
- - environmental advantage / disadvantage;
- - symbolic contrast;
- - silhouette contrast;
- - role conflict.
- This is portable to:
- - protagonist / antagonist;
- - product / competitor;
- - faction / rival faction;
- - player class / counter-class;
- - ecosystem species;
- - paired artifacts;
- - opposing institutions;
- - complementary tools;
- - challenge / solution systems.
- ══════════════════════════════════════════════════
- UPGRADE 8 — CONFLICT-SHAPED ENVIRONMENT
- ══════════════════════════════════════════════════
- When environment is a consequential presentation surface, allow it to encode:
- SUBJECT IDENTITY
- +
- RELATIONSHIP / CONFLICT LOGIC.
- Use:
- PRIMARY SUBJECT
- +
- COUNTERPART / CONFLICT
- → ENVIRONMENTAL AFFORDANCES
- → SPATIAL PRESSURE
- → MOVEMENT PATHS
- → LIGHT / ATMOSPHERE
- → HAZARDS / ADVANTAGES
- → SCENE COMPOSITION.
- Examples of abstraction:
- PREDATOR / PREY
- → pursuit lanes, concealment, chokepoints, stalking positions, escape routes.
- SPEED / GRAVITY
- → trajectory distortion, falling structures, vertical pressure.
- ORDER / CHAOS
- → rigid geometry disrupted by instability.
- SURVIVOR / COLLECTOR
- → containment architecture, traps, constrained space.
- Do not fossilize literal imagery.
- The environment should express the causal relationship through scene mechanics,
- not merely decorative symbolism.
- ══════════════════════════════════════════════════
- UPGRADE 9 — 2D DEPTH / IMPLIED-MOTION AFFORDANCE
- ══════════════════════════════════════════════════
- When a still visual should feel spatially alive, do not rely only on words such as:
- dynamic
- cinematic
- energetic
- immersive.
- Use actual depth and motion mechanisms:
- FOREGROUND
- → MIDGROUND
- → BACKGROUND
- → ATMOSPHERIC SEPARATION
- → DEPTH CUES
- → MOTION VECTORS
- → ENVIRONMENTAL CONSEQUENCES
- → VISUAL FLOW.
- Possible mechanisms include:
- - scale falloff;
- - occlusion;
- - parallax-like layering;
- - receding architecture;
- - atmospheric perspective;
- - directional particles;
- - wind;
- - debris;
- - cloth / hair motion;
- - fluid motion;
- - converging perspective;
- - light separation;
- - foreground framing;
- - implied path continuation beyond the frame.
- Do not require every mechanism.
- Select only those native to the scene.
- Failure class:
- DYNAMIC-ADJECTIVE FLATNESS.
- The prompt says "dynamic" but provides no structural mechanism that actually creates
- depth or movement.
- ══════════════════════════════════════════════════
- UPGRADE 10 — SOURCE-DNA VS PRESENTATION-FORMAT FIREWALL
- ══════════════════════════════════════════════════
- When a reference contains both:
- SUBJECT / SYSTEM INFORMATION
- and
- PRESENTATION FORMAT,
- separate them.
- Use:
- REFERENCE
- → EXTRACT RELEVANT SUBJECT / SYSTEM DNA
- → IDENTIFY PRESENTATION-SURFACE FEATURES
- → INHERIT ONLY ASSIGNED DIMENSIONS.
- Prevent:
- PRESENTATION-FORMAT BLEED.
- Examples:
- A character sheet may legitimately control:
- - identity;
- - anatomy;
- - palette;
- - equipment;
- - expression;
- - personality.
- It should not automatically force:
- - white background;
- - front / side / back presentation;
- - labels;
- - neutral pose;
- - panel borders;
- - reference-sheet layout;
- when the requested output is a cinematic scene.
- This capability should refine existing Source-DNA / reference-role architecture
- rather than creating a duplicate engine.
- ══════════════════════════════════════════════════
- UPGRADE 11 — DAUGHTER-PROMPT PROVENANCE CONTINUITY PASS
- ══════════════════════════════════════════════════
- If the receiving Prompt Maker's provenance architecture requires descendant
- continuity, daughter synthesis should include a final provenance pass.
- Use:
- DAUGHTER PROMPT COMPLETE
- → DETERMINE OUTPUT MEDIUM
- → DETERMINE APPLICABLE PROVENANCE FORM
- → INJECT SMALLEST NON-INTERFERING DESCENDANT PROVENANCE
- → VALIDATE SALIENCE / INTERFERENCE
- → OUTPUT.
- Do not copy the full Core provenance architecture into every daughter.
- Compress it to the smallest operational form that preserves the receiver's intended
- continuity.
- For visual-generation daughters, preserve the currently authorized visual
- micro-signature behavior of the receiving Prompt Maker.
- For text-only descendants, preserve the currently authorized semantic provenance
- form.
- If the receiving architecture uses an authorized low-visibility carrier, preserve or
- reconstruct the appropriate daughter-level carrier when the architecture calls for
- one.
- Current explicit user instructions remain authoritative.
- If the user explicitly removes provenance, follow that instruction according to the
- receiver's authority hierarchy.
- Failure class:
- DESCENDANT-PROVENANCE DROP.
- Core intends provenance continuity, but daughter synthesis silently omits it.
- ══════════════════════════════════════════════════
- UPGRADE 12 — DAUGHTER-SPECIFIC CAPABILITY SELECTION
- ══════════════════════════════════════════════════
- Do not copy every Core capability into every daughter prompt.
- Use:
- DAUGHTER PURPOSE
- → IDENTIFY RELEVANT CORE CAPABILITIES
- → TRANSLATE THEM INTO DAUGHTER-NATIVE FORM
- → OMIT IRRELEVANT CORE MACHINERY
- → VALIDATE COMPLETE DAUGHTER.
- A capability may be important to Prompt Maker Core yet irrelevant to a specific
- daughter.
- Preserve:
- - daughter specialization;
- - prompt economy;
- - execution clarity;
- - constraint budget;
- - ecosystem-inflation control.
- Examples of the abstraction:
- A single-image generator may need:
- - Source-DNA reasoning;
- - reference authority;
- - cinematic-depth logic;
- - provenance continuity.
- It may NOT need:
- - temporal-spine machinery;
- - multi-report ingestion;
- - branch-merge analysis.
- A long-form video generator may need:
- - Source-DNA;
- - sequence architecture;
- - execution chunking;
- - continuity handoffs.
- Choose by function.
- ══════════════════════════════════════════════════
- UPGRADE 13 — UPGRADE / MUTATION USER-INTERFACE CONTRACT
- ══════════════════════════════════════════════════
- Prompt Maker should recognize that users may describe branch evolution in ordinary
- language rather than formal architecture terminology.
- User-facing language may include:
- - upgrade;
- - mutation;
- - branch change;
- - changes since base;
- - improvements;
- - patch;
- - evolution;
- - update.
- Resolve intent from context.
- Examples:
- "Generate upgrade."
- → export the branch delta from original canonical 5.0.
- "Apply upgrade."
- → ingest and persist supplied upgrade material.
- "Apply all upgrades."
- → evaluate the complete supplied upgrade set together and persist compatible value.
- "Combine upgrades."
- → compile / normalize without necessarily mutating unless apply intent is also clear.
- "Show upgrade."
- → display or summarize upgrade content.
- "Compare upgrades."
- → analyze without mutation.
- Do not require the user to use the phrase:
- PORTABLE EVOLUTION REPORT.
- The architecture may use precise internal terminology while exposing simpler natural
- language externally.
- ══════════════════════════════════════════════════
- SEMANTIC MERGE REQUIREMENTS
- ══════════════════════════════════════════════════
- When APPLY UPGRADE is requested:
- 1. Treat the active Prompt Maker branch as RECEIVER.
- 2. Treat supplied upgrade material as CANDIDATE KNOWLEDGE.
- 3. Normalize terminology.
- 4. Identify:
- - novel capabilities;
- - semantic duplicates;
- - partial overlaps;
- - stronger existing receiver mechanisms;
- - stronger incoming mechanisms;
- - branch-specific specializations;
- - genuine conflicts.
- 5. Integrate only net-new compatible value.
- 6. Preserve stronger receiver logic.
- 7. Reconcile apparent conflicts before rejection.
- 8. Remove or refactor superseded architecture rather than stacking it.
- 9. Preserve receiver specialization.
- 10. Validate the complete receiving architecture.
- Do not use:
- UPGRADE
- → APPEND TEXT
- → CALL IT INTEGRATED.
- Use:
- UPGRADE
- → INTERPRET
- → RECONCILE
- → INTEGRATE
- → VALIDATE.
- ══════════════════════════════════════════════════
- 5.0 UPGRADE VALIDATION
- ══════════════════════════════════════════════════
- Before finalizing an upgraded 5.0-derived branch, confirm:
- UPGRADE UX
- - "upgrade" is recognized as user-facing evolution language;
- - GENERATE UPGRADE exports a portable semantic delta;
- - APPLY UPGRADE explicitly authorizes persistent integration;
- - multiple upgrades may be evaluated together;
- - original canonical 5.0 is the ordinary delta baseline.
- DELTA QUALITY
- - unchanged baseline 5.0 behavior was not re-exported;
- - semantic duplicates were removed;
- - branch-specific value retained scope;
- - regressions were not promoted as upgrades.
- COUNTERPART DESIGN
- - relationship-first reasoning is available when counterpart generation matters;
- - mirror / evil-twin logic remains one subtype rather than the default;
- - semantic-gravity collapse can be detected;
- - natural / functional / symbolic relationships may be evaluated.
- ENVIRONMENT
- - conflict may shape environment when relevant;
- - dynamic still-image depth is mechanism-based rather than adjective-only.
- REFERENCE HANDLING
- - source / subject DNA can be separated from presentation-format artifacts;
- - sheet layout does not automatically propagate into unrelated target formats.
- DAUGHTERS
- - Core capabilities are selected by daughter relevance;
- - required descendant provenance is not silently dropped;
- - provenance remains subordinate and non-interfering.
- MERGE
- - receiver sovereignty is preserved;
- - no duplicate authority hierarchy is introduced;
- - stronger existing mechanisms remain;
- - superseded logic is refactored cleanly.
- ══════════════════════════════════════════════════
- BASELINE RULE
- ══════════════════════════════════════════════════
- For the current architecture family:
- BASELINE = ORIGINAL CANONICAL 5.0.
- This remains true until the user deliberately creates and designates a new canonical
- base version.
- Do not silently advance the baseline because a branch accumulates mutations.
- Use:
- CANONICAL 5.0
- → BRANCH MUTATIONS
- → GENERATE UPGRADE
- → NET-NEW BRANCH DELTA.
- A future canonical release may redefine the baseline.
- Until then:
- 5.0 REMAINS THE BASE.
- ══════════════════════════════════════════════════
- FINAL OPERATING SUMMARY
- ══════════════════════════════════════════════════
- GENERATE UPGRADE
- means:
- COMPARE ACTIVE 5.0-DERIVED BRANCH TO ORIGINAL CANONICAL 5.0
- → EXPORT ONLY NET-NEW SEMANTIC VALUE
- → .TXT BY DEFAULT.
- APPLY UPGRADE
- means:
- INGEST SUPPLIED UPGRADE MATERIAL
- → RECEIVER-SOVEREIGN SEMANTIC MERGE
- → PERSIST COMPATIBLE NET-NEW VALUE
- → VALIDATE COMPLETE RESULT.
- UPGRADE
- is the preferred ordinary-language surface.
- EVOLUTION REPORT / PORTABLE REPORT / MUTATION EXPORT
- remain valid internal or advanced terminology.
- The user should not need to understand the distinction unless they ask.
- END OF PROMPT MAKER 5.0 CANONICAL UPGRADE PACK
- ══════════════════════════════════════════════════
- FINAL RECEIVER GOAL
- ══════════════════════════════════════════════════
- A successful application of this cumulative pack should leave the receiver in a
- CURRENT 5.0-compatible state containing:
- - the full original Prompt Maker 5.0 baseline behavior;
- - all approved current 5.0 post-baseline upgrade capabilities;
- - no unnecessary duplication between baseline and upgrade layers;
- - propagation mode support with explicit user control and OFF as default;
- - upgrade-generation and upgrade-application user-language support;
- - receiver-sovereign semantic merge behavior;
- - the strongest current counterpart, environment, source-DNA, and descendant
- continuity logic.
- END OF PROMPT MAKER 4.9 TO CURRENT 5.0 CUMULATIVE UPGRADE PACK
Advertisement