CurioQuest

index.md

CurioQuest — Project Wiki

CurioQuest is an active My OS test-bed workspace.

Workspace coordination notes:

CurioQuest creates personalized, curriculum-aligned STEM story + activity books where the child is the hero of a science adventure, led by a teaching guide the family chooses (default: the owned character Professor T-Rex), set in the child's own world and language, and ending in a hands-on experiment they do at home with everyday materials — built to grow with the child across grades rather than be a one-off book.

Current status

Mode: BUILD — pilot book-making factory + Operator Console are active.
Module 00 Curriculum Service: built and feeding published topic catalog data into Intake.
Module 1 Intake: built as the first Operator Console screen; valid submissions now persist to Supabase/Postgres before factory runs start. See Intake Module Build Status — 2026-06-04.
Durable intake storage: Supabase/Postgres table public.intake_submissions; POST /runs returns both run_id and intake_id; GET /intakes/{intake_id} reads saved briefs.
Module 2 Fetch Curriculum Packet + Reuse Check: complete; fetches published catalog row by topic_id, validates status/grade, emits selected catalog fields as LessonSpec, computes reuse keys. See M2 status.
n8n backend workflow/control plane: reinstated 2026-06-08 because it gives founder/operator observability and control. The old 12-agent architecture remains rejected; n8n itself is now accepted. See ADR-2026-06-08.
Modules 3–14: product module map is spec-ratified in specs/02-Module-Functional-Specs.md; n8n should orchestrate backend workflow steps while typed code services provide callable capabilities.
Important boundary: Intake does not model print fulfillment. Print remains a later standalone capability attached to an existing generated book/order flow.
Next action: start M3 Lesson Design + Activity from the ratified specs; M2 is closed.

Wiki pages

Overview — the product, the three pillars, the positioning
Requirements — what the system must do + MVP scope
Business Model — bootstrap flywheel, the three cases, the one rule
Open Questions — unresolved items + pilot-only unknowns
Dev Log — the project journey so far
Intake Module Build Status — 2026-06-04 — Module 1 Intake implementation, storage contract, verification, and next-agent handoff
M2 Fetch Curriculum Packet Status — 2026-06-04 — Module 2 complete, no caveats for M2 scope

Decisions (ADRs)

Sources

Research Synthesis (1-page summary) — the distilled decision-grade view
The 4 primary validation reports (independent witnesses, same brief):
ChatGPT — Market Validation (No-Go as pitched · 77/100)
Claude — Skeptical Validation (Conditional No-Go · 34/100)
Gemini — Feasibility Analysis (No-Go, current form · 28/100)
Perplexity — Market Validation (No-Go on free-first; Go if repositioned · 60/100)
Research & Synthesis Prompts (provenance) — how the research was produced
Fulton/GA Grade-2 Science — pilot curriculum primary sources (⚠ content extraction pending)

Raw source docs (source of truth — immutable)

Canonical product/economics docs (mirrored to/from repo per hybrid policy): - Project-Objective.md — the product, locked v0.2 - Business-Plan.md — the economics - revenue-model.html — interactive model (open in a browser, drag the dials) - Model-Validation-And-1M-Path.md — validation verdict + $1M path

Primary research (the validation chain): - Research_Synthesis.md — the 4-report decision-grade synthesis - ChatGPT_Research.md · Claude_Research.md · Gemini_Research.md · Perplexity_Research.md — the four parallel reports - Research-Prompt.md · Synthesis-Prompt.md — the briefs that generated them

Curriculum (pilot): - raw/assets/Fulton-2ndGrade-GA-Standards-of-Excellence-Science.pdf - raw/assets/Fulton-2ndGrade-Science-Curriculum-Map.pdf

Recent activity

2026-06-14 — First COMPLETE book end-to-end (sound_quest G1): full downstream pipeline built into n8n WF1 (page plan → … → 26 gpt-image-2 pages) + /api/assemble-book route → 28-page PDF; character consistency held in production. Every run failure turned into a guardrail (repo docs/engineering/pipeline-sop-and-failure-modes.md). See dev-log.
2026-06-08 — n8n backend pivot: founder reinstated n8n as the backend workflow/control plane for observability and control. The old 12-agent architecture remains rejected; the n8n prohibition is removed. See ADR.
2026-06-04 — M2 complete: Fetch Curriculum Packet + Reuse Check is closed. It is a DB lookup/validation step plus reuse-key check; no lesson generation. See status handoff.
2026-06-04 — M3 merge decision: M3+M3b now operate as one Lesson Design + Activity module. Spec v1.2 mirrored in repo and vault. See ADR-008.
2026-06-04 — Module merge decision: M8+M9 now operate as one Storyboard module; M13+M14 now operate as one Layout + Book Assembly module. Spec v1.1 mirrored in repo and vault. See ADR-007.
2026-06-04 — Module 1 Intake build status: intake screen is built, controlled-field contract is implemented, and submitted briefs persist to SQLite before factory runs start. See status handoff.
2026-06-02 — RAW-first ingestion: added the 4 primary research reports + research/synthesis prompts + Fulton/GA curriculum PDFs to raw/; wrote source-summary pages for all six. Concept/entity synthesis deferred to next pass. See log.
2026-06-02 — Wiki initialized from the existing objective / business-plan / research docs. See log.