# Prototype v0.3 Decision **The Curiosity Lab** Version: 0.3 Status: Accepted product decision Decision date: 2026-07-22 Path: `/mvp/Prototype-v0.3-Decision.md` --- ## DECISION **Prototype v0.3 will use the Living Narrative Canvas as the current experience model.** The previous linear prototype models are retained as research history, but they are no longer the active product direction. The current model is: ```text Persistent Object Narrative ↔ Narrative Anchors ↔ Associations ↔ Narrative Insert Preview ↔ Meaning Weave ↔ Expanded Narrative ↔ Meaning Trace ``` The core principle is: > **The narrative is not the endpoint of exploration. > The narrative is the surface on which exploration happens.** --- ## STATUS ```text Accepted Current Authoritative for v0.3 ``` This decision supersedes earlier product instructions that define the prototype primarily as: ```text a linear guided journey a sequence of discovery screens a doorway-selection flow a return-to-object loop a saved-narrative exercise ``` Those models remain useful as evidence of what was tested and learned. They must not be treated as the current experience architecture. --- ## CONTEXT The project initially followed a theory-first approach. The working process was approximately: ```text research theory → ideal discovery scenario → interface model → implementation → practical feedback ``` The first MVP deliberately kept the scope small. That was useful because it exposed a fundamental weakness before the project expanded. The prototype demonstrated that: ```text a pathway can be clear a pathway can be usable a pathway can contain relevant information and still feel lifeless ``` The experience felt: ```text dry prescribed clunky amateurish linear too much like an educational wizard ``` The problem was not merely presentation or styling. The deeper problem was the experience shape. The prototype treated meaning as something delivered through a sequence. The emerging vision treats meaning as something that accumulates around an object. --- ## WHAT WAS LEARNED The most useful outcome of the MVP was the clarification that: > **The initial description of an object should remain visible and expand with meaning through exploration of motifs, themes, context and relationships.** The user should not be moved away from the object into a chain of screens. The user should navigate around the object narrative. Each useful association should return meaning to that narrative. The result should be visible. The object description should gradually become: ```text description + basic facts + historical context + obscure references + motifs + themes + categories + structural echoes + accumulated meaning ``` The experience should therefore feel less like: ```text follow this pathway ``` and more like: ```text touch this story see what it opens add that meaning back watch the story deepen ``` --- ## EARLIER PROTOTYPE STATUS ### Prototype v0.1 Approximate model: ```text Object → Guided Observation → Question → Labelled Doorways → Explanation → Sensemaking Trail → Reflection ``` Assessment: ```text Useful as an early research instrument. Rejected as the active experience model. ``` ### Prototype v0.2 Approximate model: ```text Look → Tension → Context → Insight → Return → Doorway → Reflection → Saved Narrative ``` Assessment: ```text Improved narrative intent. Still too linear and prescribed. Rejected as the active experience model. ``` ### Prototype v0.3 Current model: ```text Living Narrative Canvas ``` Assessment: ```text Current direction. Requires practical paper testing before further broad implementation. ``` --- ## RATIONALE The Living Narrative Canvas is a better fit for the project because it supports the original research ambition more directly. The Curiosity Lab is not primarily trying to guide users through a predefined explanation. It is trying to explore whether meaningful connections can cultivate curiosity. A persistent narrative canvas supports this by allowing: ```text non-linear exploration visible meaning accretion motif and theme discovery contextual expansion unexpected but explainable connections structural echoes across disciplines return to the object without a forced return step ``` The object remains the organising centre. The user can move around it without losing the story. The system can show how the story changes through exploration. --- ## CURRENT WORKING HYPOTHESIS > **A user will experience more curiosity and meaning when a basic object description visibly expands through explored associations than when they are led through a fixed linear explanation.** Supporting hypotheses: ```text Narrative anchors will invite voluntary exploration. Narrative inserts will make meaning visible. Themes will feel more meaningful when they emerge rather than appear as a menu. Obscure references will increase worldness when they reconnect to the object. Structural Echoes will deepen understanding when the shared pattern is explained. The expanding narrative will help users see how their interpretation changed. ``` --- ## CURRENT PRODUCT PRINCIPLES ### 1. Keep the object narrative visible The user should not lose the object while exploring its wider meanings. ### 2. Meaning should accrete visibly The user should see the narrative becoming richer. ### 3. Associations need reasons A connection must explain why it exists. ### 4. Specific before abstract Begin with the object, its details and its context before moving to broader patterns. ### 5. Themes should emerge Themes should be revealed through exploration rather than presented as a topic menu. ### 6. Surprise must resolve into meaning Unexpected links should not remain clever curiosities. They must become understandable. ### 7. Structural Echoes require boundaries The system must explain: ```text what is shared why it matters where the comparison stops ``` ### 8. The user does not write the whole narrative The system constructs the expanded narrative from the associations explored and woven. Optional reflection or a next question may be captured separately. ### 9. Practical testing before expansion The interaction should work in a paper test before more implementation effort is authorised. ### 10. No broad build without visible acceptance criteria Each implementation task must produce a specific, visible and manually testable change. --- ## CURRENT AUTHORITATIVE FILES The current v0.3 direction is primarily defined by: ```text /CLAUDE.md /mvp/Prototype-v0.3-Decision.md /mvp/Living-Narrative-Canvas-Spec.md /mvp/first-journey-roman-coin.json /domains/07-Patterns-Across-Disciplines.md ``` When older files conflict with these documents, the v0.3 files take priority. --- ## STALE OR SUPERSEDED GUIDANCE The following files may contain valuable historical material, but they must not be treated as current until consolidated: ```text /mvp/Product-Vision.md /mvp/First-Journey-Roman-Coin.md /mvp/Implementation-Plan.md /mvp/MVP-Backlog.md /mvp/Experiments.md /mvp/Test-Protocol.md /mvp/Test-Note-Sheet.md /mvp/Prototype-Decisions.md /mvp/prototype/README.md /mvp/prototype/Angular-Implementation-Plan.md /mvp/prototype/Angular-v0.2-Implementation-Plan.md /architecture/Experience-Patterns.md /architecture/Pattern-Catalogue.md /architecture/Ontology.md /references/Terminology.md ``` These should be: ```text rewritten consolidated clearly marked historical or moved to an archive ``` They should not remain ambiguous sources of product authority. --- ## CURRENT IMPLEMENTATION POLICY Broad implementation is paused. The current priority is: ```text documentation consolidation content clarification paper-test preparation practical observation small implementation slices ``` Do not authorise prompts such as: ```text Build v0.3. Refactor the prototype. Implement the Living Narrative Canvas. Improve the user experience. ``` These are too broad and risk expensive but superficial work. Implementation prompts should instead be narrow. Example: ```text Keep the base narrative visible while a selected anchor reveals associations in a side panel. Do not change routing, services or unrelated components. ``` --- ## PAPER-TEST REQUIREMENT Before the next broad implementation phase, the v0.3 interaction should be tested without Angular. The test should include: ```text a visible base narrative selectable anchors association summaries relationship basis narrative insert previews an expanded narrative area a simple meaning trace ``` The paper test should answer: ```text Does the narrative become more interesting? Does the user voluntarily explore another anchor? Does an association change how the object is understood? Can the user explain why a connection mattered? Do obscure references increase worldness or distract? Do Structural Echoes clarify a pattern or feel forced? Does the accumulated narrative remain coherent? ``` --- ## ACCEPTANCE CRITERIA FOR MOVING FORWARD Further implementation may proceed when: ```text the v0.3 source of truth is unambiguous; the stale linear guidance is archived or rewritten; the Roman coin content is coherent enough to paper-test; the paper test produces at least one meaningful change in perception; the user voluntarily wants to explore another association; the narrative remains coherent after several inserts; the next implementation task can be described as one visible slice. ``` --- ## CONSEQUENCES ### Positive consequences ```text Lower implementation cost before validation. Clearer separation between research hypothesis and interface assumption. Less risk of building a polished version of the wrong interaction model. Better traceability of product learning. Stronger alignment between ontology, content and interface. More useful Claude prompts with narrower scope. ``` ### Costs and trade-offs ```text Some existing prototype work becomes historical rather than foundational. Several documents require consolidation. Implementation progress may appear slower temporarily. The project must tolerate unresolved design questions during paper testing. The Living Narrative Canvas may also fail and must remain testable rather than protected. ``` --- ## RISKS ### Risk 1 — Narrative accumulation becomes incoherent Mitigation: ```text test insert order; keep inserts short; group related meaning; limit early association count. ``` ### Risk 2 — Anchors become a disguised topic menu Mitigation: ```text integrate anchors naturally with the object narrative; avoid presenting all themes as equal categories; show why each association matters. ``` ### Risk 3 — Structural Echoes become clever but shallow Mitigation: ```text state the shared structure; state the limits; require a meaningful return to the object. ``` ### Risk 4 — Practical testing becomes detached from research Mitigation: ```text record hypotheses; capture observations; update the research register; preserve evidence and interpretation distinctions. ``` ### Risk 5 — Documentation consolidation becomes endless Mitigation: ```text update only the documents needed to run and interpret the next test; defer completeness that does not affect the current experiment. ``` --- ## REVIEW TRIGGERS Review this decision if: ```text paper tests show that users still expect a linear path; the expanding narrative becomes unreadable; anchors do not invite exploration; associations feel arbitrary; themes do not emerge naturally; users do not notice meaning accretion; a different interaction model repeatedly produces stronger curiosity; the v0.3 terminology obstructs rather than clarifies testing. ``` This decision should not be protected from evidence. The Living Narrative Canvas is the current hypothesis, not a doctrine. --- ## NEXT ACTIONS Recommended order: ```text 1. Use this file as the formal v0.3 decision record. 2. Normalise repository paths. 3. Rewrite Product-Vision.md for v0.3. 4. Consolidate Experience-Patterns.md. 5. Consolidate Pattern-Catalogue.md. 6. Consolidate Ontology.md. 7. Consolidate Terminology.md. 8. Create Roman-Coin-Living-Canvas-Content.md. 9. Rewrite the experiment and paper-test documents. 10. Conduct the paper test. 11. Authorise one small implementation slice from observed findings. ``` --- ## FINAL DECISION STATEMENT > **The Curiosity Lab will proceed with Prototype v0.3 as a Living Narrative Canvas. Earlier linear models will be retained as research history but will no longer direct the active product. Documentation consolidation and practical paper testing will precede further broad implementation.**