It’s a multi-agent web-novel writing toolkit built for tools like Claude Code , Codex, and Hermes Agent . Simply put, it lets you form a virtual novel studio with six AI agents that take a story from project planning, worldbuilding, outlines, and setting cards all the way through to the manuscript, the final edit, and cover illustrations. You only need to hand it a rough concept.

Why Build a Pipeline for Writing Fiction

A pipeline for video makes intuitive sense: video is already a divided job—research, scripting, voiceover, editing. But writing a novel sounds like one holistic act. Why break it apart?

Because I’ve written long-form with a single large model myself, and I kept hitting the same potholes:

• It drifts the longer it goes. At the start it remembers the protagonist has a scar on their left hand; by chapter thirty that scar has vanished. Settings, names, places—they quietly start contradicting each other.
• It collapses when asked to write too much at once. Tell the model to “keep going,” and it either rushes through a few paragraphs to wrap up, or it starts forgetting earlier events once the context fills up.
• A strong “AI flavor.” Just like with video scripts, models love to pile on ornate parallelism and grandiose adjectives in prose, and it reads as fake.

All of these come down to one thing: a single brain trying to manage the global setting and the line-by-line writing at the same time. So I wanted to try the same trick as the video project—split novel creation into roles that each own one slice, and have them hand off through files rather than relying on one conversation’s context to carry everything.

Six Agents, Each Owning One Slice

In this project, I broke novel creation into six roles:

• Chief Editor: The studio’s general manager. Receives your concept, drafts the implementation plan, drives the workflow, reports back in stages, and signs off on final delivery.
• Architect: Lays the foundation. Worldbuilding, the synopsis, and chapter outlines all come from here.
• Lore Master: Turns abstract settings into concrete “cards”—character cards, location cards, key-item cards, key-plot cards—a dictionary the writer can look things up in.
• Writer: Drafts the manuscript in batches.
• Editor: Polishes the prose and checks it item by item against the synopsis and the setting cards for consistency.
• Designer: Produces text-to-image prompts and specs for the cover, character art, and key-scene illustrations.

Above all of them sits a shared “constitution”—the master skill novel-studio—which defines the pipeline’s state machine, the project directory layout, the batching rules, and the quality gates. All collaboration happens through files in the project workspace, with project.yaml tracking progress, so even if you write five chapters today and pick it back up next week, it can resume from where it left off.

The Iron Rule of “Mandatory Batching”

If “de-AI polishing” was what I cared about most in the video project, then in the novel project the thing I obsessed over was mandatory batching.

A long novel easily runs to dozens or hundreds of chapters. Let the model “write it all in one go” and the second half will fall off a quality cliff. So I baked in an iron rule: by default, only 5 chapters per batch. When a batch is done, it must stop, the Chief Editor sums up that batch’s results and reports back to you, and only after you confirm does the next batch begin.

This buys two things. First, it cuts the quality risk into small chunks—if a batch isn’t good, you re-run just that batch instead of starting over. Second, it fits the reality that agents “forget”: each batch ends by producing a rolling recap and a settings index, and the next batch starts by reading that compressed memory rather than wading through tens of thousands of words of raw text.

To fight the “drifts the longer it goes” problem, I also set up three consistency gates: once a name is frozen in a setting card it can’t be changed → the writer self-checks before submitting → the editor does two passes of cross-checking. Run that gauntlet, and that scar on the protagonist’s left hand isn’t going anywhere.

A Few Thoughts

Having built the video studio and then this one, I’m increasingly convinced that what multi-agent collaboration actually solves isn’t “making AI smarter”—it’s “keeping AI from tangling itself up over a long process.”

A single model’s ceiling is genuinely high, but ask it to carry a whole novel alone and it ping-pongs between the big picture and the fine details, losing one while tending the other. Split the task apart and use files as the handoff surface, and each agent only has to focus on one small thing at a time—which, counterintuitively, makes it steadier.

Of course, it can’t replace a genuinely good story. However self-consistent the settings or however smooth the pipeline, whether it moves anyone in the end still comes down to the concept you fed it at the start, and to whether you’re willing to stop when it veers off and steer it back by hand.

If you’d like to build your own novel studio, take a look at the code and try dropping it into your Claude Code or Hermes Agent.

It’s a multi-agent video creation toolkit built for Hermes Agent . Simply put, it lets you form a virtual video studio with 7 AI agents, taking over the entire workflow—from in-depth research, scriptwriting, and TTS voiceovers to Remotion animation rendering and multi-platform SEO packaging.

Why Build This Tool

As AI capabilities continue to improve, creating short videos with AI has become incredibly easy. Nowadays, we can frequently scroll through various AI-generated videos on TikTok or Bilibili. However, whether it’s the advanced Seedance 2.0 or Gemini omni, they share common issues like slow generation speeds and difficulty in control. Sometimes, to generate a satisfactory video, we have to repeatedly roll the dice, constantly picking and splicing, which consumes a vast amount of time—let alone mass-producing short videos.

While I believe AI will eventually solve these problems given the current pace of development, these pain points still exist right now. So I started wondering: could we make AI video generation slightly more controllable through a multi-agent collaborative workflow? At the very least, when we are unsatisfied with the final video output, we wouldn’t have to start from scratch but rather tweak a specific node in the workflow, saving a considerable amount of time.

Thus, this project was born. It is particularly suited for producing batches of popular science short videos. From gathering materials, writing verbatim scripts, recording or configuring TTS, editing, and animating, to packaging titles and descriptions for various video platforms—this entire process can be collaboratively completed by AI agents with diverse characteristics. You only need to provide the topic, and AI can mass-produce videos for you like an assembly line. Of course, you can do the voiceovers yourself and add your own A-roll footage to achieve even higher video quality.

7 Agents, 6 Stages

In this project, I broke down the video production workflow into the collaboration of 7 specific roles:

• Director: The general manager of the entire studio, responsible for receiving your topics, breaking down tasks, and dispatching them to other agents.
• Researcher: Responsible for in-depth research and outputting structured research data.
• Writer: Drafts the initial video script based on the research data.
• Editor: Exclusively responsible for “de-AI polishing” and finalizing the script.
• Narrator: Calls TTS tools to generate voiceovers and outputs timeline synchronization files.
• Renderer: Transforms text and audio into dynamic video frames based on Remotion.
• Packager: Generates titles, descriptions, and tags tailored for platforms like YouTube and Bilibili.

The entire pipeline flows completely automatically. You can choose different working modes: if you want to chime in at any time, you can pull them into a “group chat” and watch them discuss; if you just want to see the results, you can contact the Director one-on-one and let it “delegate” work in the background; if you need mass production, you can also manage the progress using a Kanban mode.

The Necessity of “De-AI Polishing”

Throughout the entire pipeline, what I care about the most—and spent the most effort tuning—is the Editor stage.

Current large models share a common flaw when writing articles: they love using overly enthusiastic marketing tones filled with grandiose adjectives. This feels disconnected from our daily reading experience. I personally have always preferred restrained, natural, and clear writing. If I were to read the script myself, that strong “AI flavor” would make me feel very uncomfortable.

Therefore, after the Writer produces the first draft, I forcibly introduced the Editor role. Its task is to perform “de-AI polishing”: removing pompous vocabulary, breaking long sentences into short ones, eliminating stereotypical patterns, and making the script return to the natural tone of human speech as much as possible.

A Few Thoughts

After completing this project, I have a very strong feeling: multi-agent collaboration might truly change the survival state of individual creators.

In the past, when we talked about a “one-person studio,” it often carried a sense of ascetic tragedy. But now, if you know how to orchestrate AI, you can genuinely have your own virtual employees. In Video Studio, you only need to decide on a good topic. AI will automatically run through the rest of the dirty work for you.

Technology is leveling the execution threshold at an unprecedented speed. When tools are no longer an obstacle, what ultimately widens the gap might once again return to those things in our minds that truly belong to humans: your taste, your restraint, and your unique understanding of the world.

If you are interested, feel free to check out the code and try deploying it into your Hermes Agent.