The prior commit's git add silently dropped these five files because one path in the same invocation didn't exist. Docs, solution file, and example config still needed the paired update: Novelly.slnx drops the Novelly.Mcp project entry, README and CLAUDE.md describe MCP as an in-API HTTP endpoint instead of a stdio binary, .mcp.json.example uses the type: http shape, and outline-importer.md's tool references are fixed to real tool names.
13 KiB
Novelly
Software for planning and writing a novel. You outline the book, keep character dossiers, break chapters into scenes, and draft prose — with a Claude-powered agent embedded in the app that can read and edit the same data you can, and an MCP endpoint that exposes that same data to Claude Code, Claude Desktop, or any other MCP client.
The point of the three-way arrangement is that there is exactly one source of truth. The React UI's REST calls, the embedded agent, and MCP clients all resolve to the same application services in-process, so an edit made from a chat in Claude Code and an edit made by typing in the browser are the same edit.
Stack
| Piece | Built with |
|---|---|
Novelly.Api |
ASP.NET Core 10 minimal APIs, EF Core 10 + SQLite, Anthropic SDK, OpenAPI, MCP over Streamable HTTP (ModelContextProtocol.AspNetCore) |
Novelly.AppHost |
.NET Aspire orchestration for the API and the web client |
Novelly.ServiceDefaults |
Shared OpenTelemetry, health checks and service discovery |
Novelly.Web |
React 19, TypeScript, Vite, TanStack Query, Tailwind v4 |
The back end is one project organised by feature, not by layer. Each feature folder —
Novels/, Characters/, Chapters/, Beats/, Tags/, Locations/, Agent/ — holds its
entity, DTOs, service and endpoints together, so adding a capability means touching one
folder rather than four. Common/ holds what genuinely crosses features and Data/ holds
the DbContext and migrations.
Running it
Prerequisites: .NET 10 SDK and Node 20+.
Everything at once, through Aspire:
cd src/Novelly.Web && npm install && cd -
dotnet run --project src/Novelly.AppHost
That starts the API on :5080 and the Vite dev server on :5173, and opens the Aspire dashboard for logs, traces and metrics across both.
Or run the two halves separately:
# 1. API — creates and migrates novel.db on first run, listens on :5080
ASPNETCORE_URLS=http://localhost:5080 dotnet run --project src/Novelly.Api
# 2. Web — dev server on :5173, proxies /api to :5080
cd src/Novelly.Web && npm install && npm run dev
Open http://localhost:5173.
The app is fully usable without an Anthropic key — only the Agent tab needs one. To turn the agent on:
export ANTHROPIC_API_KEY=sk-ant-...
Without it, agent endpoints return 503 Agent unavailable with an explanatory message
and everything else keeps working.
Tests
dotnet test # 174 tests
cd src/Novelly.Web && npm run build # typecheck + bundle
git push runs both through a Husky pre-push hook (scripts/ci/prepush.sh). Install the
hook once with npm install at the repo root.
Tests run against real in-memory SQLite rather than the EF in-memory provider, so they exercise the cascade deletes and query translation the app actually ships with.
Configuration
src/Novelly.Api/appsettings.json:
{
"ConnectionStrings": { "Novel": "Data Source=novel.db" },
"Cors": { "Origins": [ "http://localhost:5173" ] },
"Agent": {
"Model": "claude-opus-5",
"MaxTokens": 16000,
"Effort": "high", // low | medium | high | max
"MaxIterations": 12 // tool-call ceiling per user turn
}
}
The API key is read from ANTHROPIC_API_KEY or, if you prefer, Agent:ApiKey — keep it
out of appsettings.json and use user-secrets or the environment.
The data model
Novel ──┬── Character ──┬── CharacterRelationship
│ └── CharacterArcStage (the arc: flat, ordered)
├── Chapter ──── Beat (the outline: flat, ordered)
├── Tag (applied to characters, chapters and beats)
├── Location (applied to chapters)
├── OpenQuestion (attached to a chapter and/or a character)
└── AgentConversation ── AgentMessage
A chapter outline is a paragraph plus a table. The paragraph is the chapter's
Summary; the table is its beats. Each beat is one row:
| Column | What goes in it |
|---|---|
| Beat | A three-to-five word handle — "she burns the atlas", not a sentence |
| Characters | Who's in the beat. Optional; not every beat belongs to anyone, and a beat can name more than one |
| What happened | The event itself |
| What's next | What it sets in motion — the hook into the following beat |
Beats are flat and ordered by SortOrder within their chapter. There is no nesting and
no tree — reordering is one call that takes the beat ids in the order wanted. The outline
tab also rolls up the distinct set of characters across a chapter's beats under the beat
and word counts, each linking to that character's page — a quick cast list without
opening every beat.
Beats plan; the chapter's Prose carries the draft. A chapter has one prose field,
written and redrafted in place; WordCount is recomputed from it whenever it changes.
There is no separate scene entity — the outline (beats) and the draft (prose) are the
two views of the same chapter.
Main characters carry the book; supporting characters hold it up. A character's
Importance (Main or Supporting) is separate from their Role — role is the part they
play in the story, importance is how much of it they take, and a mentor can be either.
Characters start supporting and get promoted. The two together drive the ordering, so the
protagonist is always the first name on the list.
A main character's arc is a table, not a paragraph. ArcSummary still holds the
sentence version; CharacterArcStage breaks the same change into ordered steps, each
optionally pinned to the chapter it lands in. It is the same flat-and-ordered shape as the
beat table, for the same reason. Nothing refuses an arc on a supporting character —
demoting someone should not delete their work.
The character page reads the outlines back. GET /api/characters/{id}/beats returns
every beat a character appears in, in manuscript order, each carrying its chapter so the
page links straight into that chapter's outline. It is the dossier's reality check: what
they actually do on the page, as against what the sheet claims about them.
Open questions are what you have not decided. A question hangs off a chapter outline, a character, both, or neither. It can be resolved, reopened or deleted, and resolving can append the decision to the notes of whatever it was attached to — so a settled question ends up where you re-read it rather than in a list you have stopped looking at. Resolved questions drop off the list unless you ask for them.
Tags cross-reference the book. A tag is scoped to one novel, unique by name
(case-insensitively), and can be attached to any character, chapter or beat. Applying an
unknown tag by name creates it, so tagging is one action rather than two. GET /api/tags/{id}/references returns everything carrying a tag, which is how you trace a
motif or a thread across all three kinds at once.
A chapter can have several locations. Locations replaced the old single free-text
Setting field: a chapter now carries a multi-select list of locations (where and when
it takes place), each a novel-scoped, name-deduped entity — applying an unknown location
by name creates it, same as tags. A "Locations" tab on the novel lists every location
with its chapter count, and GET /api/locations/{id}/references lists every chapter set
there.
The embedded agent
NovelAgentService runs the tool-use loop: it calls the Messages API, executes any tools
Claude asks for, feeds every result back in a single user turn, and repeats until Claude
stops asking. It has 33 tools covering the brief, characters and their arcs, chapter
outlines (beats), tags, locations and open questions — all of them going through the same
application services the REST API uses.
A few deliberate choices worth knowing about:
- Conversation history replays as text only. Tool calls are not replayed into the transcript. The agent re-reads current state through its tools instead, which is more reliable than trusting a record of edits that may since have changed in the UI.
- The user's turn is persisted before the loop runs, so a question is recorded even if the model call fails.
- Tool failures come back as
is_errorresults, not exceptions — the model reads the message and corrects itself. MaxIterationscaps tool calls per turn. On hitting it the agent says so rather than silently truncating.- The system prompt is cached (
cache_control: ephemeral), so every turn after the first reads it back at a fraction of the input price.
The MCP server
The API itself serves MCP over Streamable HTTP at POST /mcp, exposing 45 tools that call
the same application services the REST endpoints and the embedded web agent call — it holds
no domain logic of its own, and there's nothing to build or publish separately. The API
process just needs to be running; there's no separate subprocess to keep in sync with it.
Copy .mcp.json.example to .mcp.json (gitignored, since it carries your API key) and
fill in the key:
{
"mcpServers": {
"novelly": {
"type": "http",
"url": "http://localhost:5080/mcp",
"headers": {
"X-Novelly-Api-Key": "<matches the API's Auth:ServiceApiKey user secret>"
}
}
}
}
The API must be running, with Auth:ServiceApiKey set (e.g. via
dotnet user-secrets set Auth:ServiceApiKey <key> -p src/Novelly.Api) to the same value
as the X-Novelly-Api-Key header above. If the API is not running, or the key is missing or
mismatched, the request 401s.
Tool argument names are camelCase, matching the REST API and every other MCP argument name
this project has ever used. create_novel called over MCP is owned by the seeded service
user (an Admin), not whichever person is signed into the web app.
Importing an existing outline
.claude/agents/outline-importer.md is a Claude Code subagent that reads an author's outline
already on disk — chapter files and character dossiers in the folder shape described in that
file — and writes it into a new Novelly project over the MCP server above. It's read-only against
the source files and keeps a resumable ledger (.novelly-import.json) next to them, since a full
book is more tool calls than fit in one turn. Invoke it with the source folder path once the API is
running and .mcp.json is set up.
API surface
GET /api/health, plus:
| Resource | Routes |
|---|---|
| Novels | GET|POST /api/novels, GET|PATCH|DELETE /api/novels/{id} |
| Characters | GET|POST /api/novels/{id}/characters, GET|PATCH|DELETE /api/characters/{id}, POST /api/characters/{id}/relationships |
| Chapters | GET|POST /api/novels/{id}/chapters, GET|PATCH|DELETE /api/chapters/{id} |
| Beats | GET|POST /api/chapters/{id}/beats, POST /api/chapters/{id}/beats/reorder, GET|PATCH|DELETE /api/beats/{id} |
| Tags | GET|POST /api/novels/{id}/tags, GET /api/tags/{id}/references, PATCH|DELETE /api/tags/{id} |
| Locations | GET|POST /api/novels/{id}/locations, GET /api/locations/{id}/references, PATCH|DELETE /api/locations/{id} |
| Arcs | GET|POST /api/characters/{id}/arc, POST /api/characters/{id}/arc/reorder, GET|PATCH|DELETE /api/arc-stages/{id} |
| Questions | GET|POST /api/novels/{id}/questions, GET|PATCH|DELETE /api/questions/{id}, POST /api/questions/{id}/resolve, POST /api/questions/{id}/reopen |
| Agent | GET /api/novels/{id}/agent/conversations, POST /api/novels/{id}/agent/messages, GET|DELETE /api/conversations/{id} |
PATCH bodies are partial: an omitted field is left alone, an empty string clears it. A
tags array replaces that item's tags outright and creates any names the novel has not
seen; the same is true of a chapter's locations array. Omitting either leaves it
untouched.
Enums travel as names ("Protagonist", "Drafted"), never ordinals. In development the
OpenAPI document is at /openapi/v1.json.
Known issues
react-router-dom7.18.2 carries GHSA-qwww-vcr4-c8h2 (CSRF bypass in RSC mode). No patched release exists yet, and every version below the affected range carries 14 worse advisories. This app is a client-only SPA and does not use RSC mode, so the advisory does not apply — butnpm auditwill flag it until a fix ships. Upgrade when one does.
Where this could go next
The vertical slice is complete but thin in places. The obvious next steps:
- Stream agent responses over SSE instead of returning the finished turn.
- Drag-and-drop beat reordering (the reorder endpoint is already there; the UI uses up/down buttons).
- Filter chapters and characters by tag from the list views, not just the Tags tab.
- A manuscript export (Markdown, DOCX) built from chapters and scenes in order.
- Revision history for scene prose.
- Authentication, if this is ever going to run anywhere but localhost.