grettirjacobs/editorial-council — explained in plain English
Analysis updated 2026-05-18
Feed a full manuscript into Council to get one consolidated editorial brief covering continuity, style, and pacing.
Run only one editorial dimension, like theme or character analysis, to check a single revision quickly.
Try the tool on the included public domain novels before uploading your own manuscript.
Use local-only mode to analyze a manuscript with nothing leaving your computer.
| grettirjacobs/editorial-council | 0xallam/my-recipe | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | Python | Python | Python |
| Last pushed | — | 2022-11-22 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 2/5 | 4/5 |
| Audience | writer | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Needs a local Ollama model server or an API key, worldbuilding specifically requires a cloud model.
Council is a multi-agent editorial system built for novelists. You feed it a manuscript, and a set of specialized software agents, called clerks, each look at one aspect of the writing: continuity, outline, style, voice, characters, themes, plot, worldbuilding, and pacing. Their separate findings are combined into one editorial brief plus per-dimension structured data files. The tool is designed to run on your own computer. If you connect it to a local model server called Ollama, you can analyze a full manuscript for free, and nothing about your writing leaves your machine. If you want a stronger pass, you can instead supply your own API key, for example from MiniMax, and let a cloud model do the work. To try it out, the project comes with several public domain novels already included, such as The Picture of Dorian Gray and The Call of the Wild, so you can see how it works before pointing it at your own manuscript. Running the pipeline produces a finished brief you can read directly, and you can rerun just one part of the analysis, like the theme pass, instead of redoing everything. One part of the system, the worldbuilding analysis that extracts characters and settings into a map, needs a cloud based model rather than a local one, though a free OpenRouter key can cover that step at no cost. Every run ends with an honest result label, either complete, degraded with a note about what is missing, or failed, so you always know what you actually got. The project is written in Python and is licensed under MIT, meaning you are free to use, modify, and share it. There is also a strict privacy mode that blocks any outgoing network call except to your own machine, for anyone who wants to guarantee nothing leaves their computer.
An open-source multi-agent system that runs on your own computer to produce a full editorial brief for a novel manuscript.
Mainly Python. The stack also includes Python, Ollama, OpenRouter.
MIT license lets you use, modify, and share the code freely, including for commercial purposes.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly writer.
This repo across BitVibe Labs
double-check against the repo, no cap.