[WIP] Implement full hands-off vibe coding upgrades for E-Crafter

[Copilot]

• Explore codebase and understand existing structure
• Fix TitleBar.tsx build failure (corrupted file with embedded imports in JSX)
• Create services/llmOrchestrator.ts - Central LLM layer: prompt compaction, token budgeting, seeded RNG, IndexedDB cache, telemetry, jittered retry
• Create services/memoryService.ts - Micro-summary + continuity facts delta extraction
• Create services/loreEngine.ts - LoreStore with IndexedDB, entity extraction
• Create services/sceneChunkedGenerator.ts - Scene-chunked chapter generation (11-14 scenes, deterministic per seed)
• Create agent modules: agents/TrendSeoAgent.ts, agents/CriticAgent.ts, agents/EmotionalEditorAgent.ts, agents/FineTunerAgent.ts, agents/ContinuityAgent.ts
• Update types.ts - Add GenerationSettings, TelemetryEntry types; remove duplicate BotUpdate entry
• Create components/GenerationSettingsPanel.tsx - strategy toggle, scene range, polish toggle, expansion cap, re-roll button
• Create components/TokenTelemetryDashboard.tsx - totals per project, cache hit rate
• Update App.tsx - wire new settings panel and telemetry dashboard; add rerollNonce state
• Add tests: tests/llmOrchestrator.test.ts - 52 tests passing
• Resolve merge conflict with main (jspdf 4.2.1 CVE fix, tokenOptimizer from main, dedup BotUpdate union type)

Original prompt

Implement the full hands-off “vibe coding” upgrades for Binary Books E-Crafter in crazyrob425/KDP-E-Book-Generator, focusing on elite multi-agent orchestration with extreme token efficiency and high-quality output.

Scope (must be production-ready, no placeholders):

A) Token Efficiency Core (must be applied everywhere)

1. Create a central LLM orchestration layer used by all AI calls:
   • Prompt compaction and deduplication (canonical instruction blocks)
   • Deterministic token budgeting per task with char-based estimation and guardrails
   • Deterministic seeded randomness support for controllable variation (used for scene counts)
   • Response caching persisted in IndexedDB (and optionally file cache in Electron) keyed by normalized prompt + model + params + schema
   • Telemetry: estimated in/out tokens, cache hits, per-step totals
   • Robust retry with jittered exponential backoff
2. Add “memory without tokens” mechanism:
   • After each scene/chapter: micro-summary + continuity facts delta extraction
   • Store summaries/facts in project state (IndexedDB) and reuse them instead of resending full text

B) Default Chapter Generation: Scene-chunked with randomized scene count

1. Make scene-chunked generation the default.
2. Scene count randomized between 11 and 14 inclusive, but deterministic per (projectSeed + chapterNumber + outlineHash + rerollNonce).
3. Implement:
   • Scene planning step (schema-first JSON)
   • Per-scene drafting step with rolling micro-summary and minimal context
   • Local stitcher
   • Final polish pass that enhances emotion, improves flow, removes repetition, preserves plot facts; cap expansion at configurable 5–8% (default 6%).

C) Multi-agent swarm ecosystem (lightweight but real)
Implement agent modules that operate on compact representations and directives:

• Trend & SEO Agent: integrates existing realMarketService + Gemini to produce evidence-backed keywords/title/subtitle/blurb/category suggestions.
• Critic Agent: returns compact directives JSON (issues, severity, target scene/section, fixes).
• Emotional Editor Agent: directives to heighten emotional impact.
• Fine-Tuner/Copywriter Agent: grammar, tone, rhythm, consistency, genre conventions.
• Continuity Agent: checks against Lore Engine facts, flags conflicts, proposes edits.

D) Lore Engine (persistent world-building memory)

1. Implement LoreStore in IndexedDB with stable IDs for characters/locations/factions/rules/timeline.
2. Entity extraction from outline + scenes to update lore incrementally.
3. Continuity checking in generation loop using facts by ID.

E) UI/UX & Persistence

1. Add Generation Settings panel:
   • strategy toggle (chunked vs single-pass) default chunked
   • scene range min/max default 11–14
   • emotional polish toggle default ON
   • expansion cap default 6%
   • re-roll scene split button (increments rerollNonce per chapter)
2. Persist settings + seeds in project state with safe migration for existing saved projects.
3. Add token telemetry dashboard: totals per project, per chapter, cache hit rate.

F) Tests
Add unit tests for:

• prompt compaction
• token estimation
• deterministic RNG/seeded scene count
• cache key stability
• chunking budget guard logic

Constraints:

• Zero budget: no paid services.
• Maintain existing app flows; do not break current steps.
• Keep code fault-tolerant with clear error messaging.

Deliverable: Open a PR with all changes.

The following is the prior conversation context from the user's chat exploration (may be truncated):

User: analyze the codes line by line and create a detailed blueprint and todo

Assistant: Below is a codebase analysis + blueprint + TODO plan for crazyrob425/KDP-E-Book-Generator based on the files I could pull from the repo (root, services/, server/, electron/, hooks/, plus directory listings under components/). Where I say “line-by-line”, I’m going file-by-file and section-by-section (the repo has a very large App.tsx, so I’m describing it in contiguous “blocks” of lines / responsibilities rather than literally enumerating 45k lines one by one).

---

0) High-level: what this repo currently is

Current architecture (as implemented):

• React + Vite + TypeScript frontend (mostly driven by a single massive App.tsx orchestrator).
• AI generation via Google Gemini (client-side calls through @google/genai) in services/geminiService.ts.
• Local persistence:
  • “V5 DB” using IndexedDB wrapper in services/storageService.ts.
  • Extra “AutoSave” feature saving a JSON file via Electron IPC if running as desktop (hooks/useAutoSave.ts).
• Electron wrapper:
  • electron/main.ts sets up a frameless window + IPC handlers.
  • Bridges to “bot automation” and “market research workers”.
• Optional Node server:
  • server/server.ts provides WebSocket-based automation sessions.
  • server/automation-worker.ts Playwright automation for KDP upload flow (with CAPTCHA yield/continue).
  • ...

This pull request was created from Copilot chat.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[crazyrob425]

@copilot u rock

[crazyrob425]

@copilot resolve conflict

[Copilot]

@copilot resolve conflict

Conflict resolved in commit 3ff826d. The merge brought in the jspdf 4.2.1 CVE fix, tokenOptimizer.ts, and the updated geminiService.ts from main, while preserving all vibe-coding upgrades from this PR. Build and all 52 tests pass.