Skip to content
Back to search

draftboard

MarketplaceWorks with Claude CodeGitHub

Draftboard: a story doctor for your draft.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/kinetiq--draftboard/badge.svg)](/listings/kinetiq--draftboard)

HTML

<a href="/listings/kinetiq--draftboard"><img src="/api/listings/kinetiq--draftboard/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

In Claude Code:

/plugin marketplace add kinetiq/draftboard
/plugin install draftboard@draftboard

You also need **Python 3.10 or newer** on your PATH (`python`, `python3` or `py`). Draftboard uses only the standard library.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 12 lines
  1. {
  2. "name": "draftboard",
  3. "description": "Draftboard: a story doctor for your draft.",
  4. "owner": { "name": "kinetiq" },
  5. "plugins": [
  6. {
  7. "name": "draftboard",
  8. "source": "./",
  9. "description": "A story doctor for your draft. Diagnoses the structure; never touches the manuscript."
  10. }
  11. ]
  12. }
README.md · README · 90 lines
  1. # Draftboard
  2. **A story doctor for your draft. Diagnoses the structure; never touches the manuscript.**
  3. Draftboard is a plugin for [Claude Code](https://claude.com/claude-code). Point it at a **finished draft**, a novel or a screenplay, and it reads the whole thing through six craft frameworks at once:
  4. | Framework | Asks |
  5. |---|---|
  6. | Save the Cat beat sheet | Are the big beats there, and in the right place? |
  7. | Character arcs (K.M. Weiland) | What does each main character believe, and when does it break? |
  8. | Promises and payoffs (Brandon Sanderson) | What did the story promise, and did it pay? |
  9. | MICE threads (Mary Robinette Kowal) | Do threads close in the right order? |
  10. | Value shifts (Story Grid) | Does every chapter or scene turn something? Where does it sag? |
  11. | Genre contract | Does it deliver what its genre promises? |
  12. Then it shows you **where the frameworks agree**, because a problem that three frameworks find independently is the one to fix first.
  13. You get:
  14. - **An interactive map of your whole story.** A tension curve with a reference shape, every framework's beats, arcs and threads laid along the same timeline, an overlap view of where they agree, and every finding with its evidence and a suggested fix.
  15. - **A brief for every chapter or scene.** What to fix here, what only shows here, tension, value shifts, threads in play, and a **My notes** section for your own issues that survives every rebuild. Claude sees the brief automatically when you work on that chapter.
  16. It never edits your manuscript. Every finding is a reading through a framework, checked against quotes from your text. You decide what to do with it.
  17. ## Install
  18. In Claude Code:
  19. ```
  20. /plugin marketplace add kinetiq/draftboard
  21. /plugin install draftboard@draftboard
  22. ```
  23. You also need **Python 3.10 or newer** on your PATH (`python`, `python3` or `py`). Draftboard uses only the standard library.
  24. ## Use
  25. 1. **Set up:** `/draftboard:setup`. It looks at your project, asks a few questions (form, which files, protagonist, genre), and writes `draftboard.json`. A few minutes.
  26. 2. **Map the draft:** `/draftboard:map`. Several agents read the whole work in parallel (a 170k-word novel uses about eleven).
  27. 3. **Work chapter by chapter:** ask "what issues am I in?" while working on a chapter or scene (`/draftboard:issues`), or say "add a note to this chapter: …".
  28. ## How much it uses
  29. Each run starts about eight to eleven agents. Most of the cost is reading your text, so it grows with length. As a guide:
  30. | Work | Tokens, roughly |
  31. |---|---|
  32. | A short script or story | 350,000 |
  33. | A feature screenplay (about 25,000 words) | 530,000 |
  34. | A 90,000-word novel | 1,000,000 |
  35. | A 170,000-word novel | 1,600,000 |
  36. On larger plans that is comfortable; on smaller plans a long novel can use a big share of your limits, so run it when you won't need Claude for a while. Draftboard tells you its estimate and asks before it starts.
  37. **Use a capable model.** The analysis is careful reading and judgment; the latest Opus at medium effort works well.
  38. ## What it reads
  39. | Your project | In `draftboard.json` |
  40. |---|---|
  41. | A folder of chapter files (`01-Title.md`, `Chapter 12.txt`) | `"source": { "type": "files", "dir": "manuscript" }` |
  42. | A Fountain screenplay | `"source": { "type": "fountain", "file": "script.fountain" }` |
  43. | A Final Draft file | `"source": { "type": "fdx", "file": "script.fdx" }` |
  44. | One file per scene | `"type": "files"` with that folder |
  45. Several books, scripts or episodes can live in one project as separate volumes. Setup handles all of this; [references/config.md](references/config.md) has the details.
  46. ## Using Cursor too
  47. If your project has a `.cursor/` folder, setup offers to add a Cursor rule, so Cursor's agent also reads each chapter's brief when you work on it.
  48. ## How it works
  49. - **Readers** read the draft once, in parallel chunks, and write a shared ledger: scenes and their value shifts, threads, promises, character beats, quotes with line numbers.
  50. - **One tension agent** scores every chapter or scene on a fixed 0–10 scale, both the last page and the peak, so the scale is consistent across the whole work.
  51. - **Six framework agents** each read the ledger through their framework and file findings, with evidence quotes and where the fix belongs.
  52. - **The composer** (your Claude Code session) checks every important quote against your text, drops what doesn't hold, groups findings that share a cause, and builds the map and the briefs.
  53. The analysis lives in your project (`draftboard/` by default), so you can commit it, compare runs, and keep your notes with your draft.
  54. ## License
  55. Draftboard is source-available under the [PolyForm Shield License 1.0.0](LICENSE.md).
  56. In plain words:
  57. - **Use it for anything, including your professional writing.** Analyze books and scripts you sell, use it for clients, change it for your own needs.
  58. - **Don't build a competing product with it.** You may not ship Draftboard, or a product built on it, that competes with Draftboard: for example a writing app, AI service, or plugin that offers its analysis as a feature.
  59. The license text is the binding version; this summary is just a guide.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.