ebook

name	ebook
description	AI Ebook Generator — create, generate, research, and export ebooks
level	2

AI Ebook Generator Skill

You have access to the ebook-generator MCP server which exposes 8 tools for managing the full ebook creation pipeline.

Available MCP Tools

Tool	Description	Required params
ebook_list_projects	List all ebook projects with status	limit (optional, default 20)
ebook_create_project	Create a new ebook project	title, idea
ebook_generate	Start background generation for a project	project_id
ebook_get_status	Get generation status and progress	project_id
ebook_get_export_info	Get strategy, marketing kit, keywords, ad hooks, pricing, word count	project_id
ebook_research_market	Search Google Books + Open Library for competing ebooks	query
ebook_list_files	List all files in a project directory with sizes	project_id
ebook_read_file	Read a project file (markdown, JSON, text)	project_id, filename

ebook_create_project optional params

• product_mode: lead_magnet | paid_ebook | bonus_content | authority (default: lead_magnet)
• target_language: BCP-47 code, e.g. en, id, es (default: en)
• chapter_count: integer (default: 5)

ebook_read_file filename examples

• strategy.json, outline.json, toc.md, manuscript.md
• chapters/1.md, chapters/2.md
• cover/prompt.txt
• marketing_kit.json

---

Typical Workflows

1. Research → Create → Generate → Monitor → Export → Push to AdForge

1. Research the market
   ebook_research_market(query="productivity habits", language="en", max_results=10)

2. Create a project
   ebook_create_project(title="Atomic Habits for Entrepreneurs", idea="...", product_mode="paid_ebook", chapter_count=7)
   → returns project with id (e.g. 3)

3. Start generation
   ebook_generate(project_id=3)

4. Poll status until completed
   ebook_get_status(project_id=3)
   → check db_status and progress fields; repeat every 30–60s until status="completed"

5. Get export metadata
   ebook_get_export_info(project_id=3)
   → returns keywords, ad_hooks, suggested_price, word_count, audience, tone

6. Read generated content (optional)
   ebook_read_file(project_id=3, filename="manuscript.md")
   ebook_read_file(project_id=3, filename="strategy.json")

7. Push to AdForge landing page
   POST http://localhost:3000/api/landing
   {
     "name": "ebook-{project_id}",
     "theme": "dark",
     "product_name": "<title>",
     "price": "<suggested_price from export>",
     "pain_points": ["...", "..."],
     "benefits": ["...", "..."],
     "cta_primary": "Download Now",
     "cta_secondary": "Learn More"
   }

2. List and Inspect Existing Projects

ebook_list_projects(limit=10)
ebook_get_status(project_id=N)
ebook_list_files(project_id=N)
ebook_read_file(project_id=N, filename="outline.json")

3. Multi-language Ebook

ebook_create_project(
  title="...",
  idea="...",
  target_language="id",  # Indonesian
  product_mode="lead_magnet"
)

---

Project Status Values

Status	Meaning
draft	Created, not yet generating
generating	Pipeline running
completed	All pipeline stages done
failed	Pipeline error (check message field)

Progress tracking: ebook_get_status returns progress (0–100) and message alongside db_status.

---

Project File Layout

projects/{id}/
  strategy.json        ← audience, tone, goal
  outline.json         ← chapter list with word count targets
  toc.md               ← table of contents markdown
  manuscript.md        ← full assembled ebook
  manuscript.json      ← per-chapter word counts
  marketing_kit.json   ← keywords, ad_hooks, pricing, book_description
  qa_report.json       ← QA checks
  chapters/{n}.md      ← individual chapter files
  cover/cover.png      ← generated cover image
  cover/prompt.txt     ← AI prompt used for cover
  exports/ebook.docx   ← Word export
  exports/ebook.pdf    ← PDF export

---

Service Endpoints

• Streamlit UI: http://ebook.aitradepulse.com (port 8501)
• REST API: http://localhost:8765
  • GET /health
  • GET /api/projects
  • POST /api/projects (requires X-Api-Key header)
  • GET /api/projects/{id}
  • POST /api/projects/{id}/generate
  • GET /api/projects/{id}/status
  • GET /api/projects/{id}/export
  • GET /api/projects/{id}/download/{fmt} — fmt: docx, pdf, epub

---

AdForge Integration

After generation, push the landing page to AdForge:

POST http://localhost:3000/api/landing
Content-Type: application/json

{
  "name": "ebook-{project_id}",
  "theme": "dark",
  "product_name": "{title}",
  "price": "{suggested_price}",
  "pain_points": ["{ad_hooks[0]}", "{ad_hooks[1]}"],
  "benefits": ["{keywords derived}"],
  "cta_primary": "Download Now",
  "cta_secondary": "Learn More"
}

Use ebook_get_export_info to populate price, pain_points (from ad_hooks), and benefits (from keywords).

---

Notes

• Generation is async — always poll ebook_get_status after calling ebook_generate.
• ebook_read_file supports .md, .json, .txt, .csv, .html only.
• cover_image_base64 in export info is a PNG encoded as base64 — render inline or save to file.
• The MCP server runs in-process (no HTTP hop) — calls are direct Python module invocations.