DeepDoc
v3.0 STABLE · source-cited docs · AI chatbot

Engineering docs
your team will
actually read.

DeepDoc generates internal documentation from your codebase — architecture, endpoints, integrations, runtime surfaces — and ships an embedded chatbot that answers from real source with exact file paths.

View on GitHub
See the generated site and chatbot
47
docs pages planned from source
1,247
chatbot chunks indexed locally
3
query modes for answers

Python 3.10+ · MIT license · Any LLM provider

deepdoc.local/docs/contextly-platform-overview
Generated DeepDoc documentation site for the Contextly platform, showing structured navigation and source-grounded content Generated DeepDoc documentation site for the Contextly platform, showing structured navigation and source-grounded content
Generated nav 14 repo-aware sections
DeepDoc Chatbot
You
Where does billing verification happen?
DeepDoc
Payment verification is handled in the Razorpay billing flow, then persisted through the usage and project records.
billing/routes.py:41-88services/usage.py:12-57
Works with Python TypeScript JavaScript Go PHP Vue OpenAPI
Product proof

A complete docs site, not a markdown dump.

DeepDoc plans the sidebar, writes grounded pages, adds source-file disclosures, stages API reference pages, and embeds a chatbot that can answer across the same indexed codebase.

generated MkDocs Material site
Generated documentation page with sidebar navigation, table of contents, and source-grounded overview content Generated documentation page with sidebar navigation, table of contents, and source-grounded overview content
Chatbot included

Ask the docs like a teammate.

The assistant uses generated docs for orientation and source archives for proof, so answers come back with real paths and line ranges.

How does the embedded widget reach the backend?
It uses the widget token flow, then sends chat requests through the API client. widget/client.ts:18-64 api/chat.py:33-91
What ships
  • Structured docs nav from the planner
  • Mermaid diagrams and API reference pages
  • Incremental updates from git diffs
  • Local chatbot index and source archive
See it in action

Watch the pipeline run.

43 sec · all five phases
What it does

One command. A docs site
your team won't ignore.

01

Bucket-based generation

Architecture, endpoints, integrations, and schemas planned as cohesive docs — not one noisy page per file. The planner classifies your repo and assigns files to a reader-first structure.

auth.py, middleware.ts → Auth & Permissions
billing/routes.py, usage.py → Billing & Usage
api/webhooks.ts, stripe.py → Integrations
db/models.py, migrations/*.sql → Data & Schema
02

Evidence-grounded chatbot

An embedded assistant ships with the docs site. Answers come with exact file paths and line ranges — not hallucinations.

Where does billing verification happen?
Payment verification runs in the Razorpay billing flow, persisted through usage records.
↗ billing/routes.py:41-88
↗ services/usage.py:12-57
How does the widget token flow work?
The widget fetches a short-lived token from /api/widget-token, then attaches it to every chat request.
↗ api/widget.ts:18-64
03

Incremental updates

deepdoc update diffs against the last synced commit and regenerates only affected buckets — smart enough to replan when new integrations appear.

deepdoc update — 3 files changed
billing/routes.py regenerate
services/stripe.py regenerate
db/models.py skip
Pipeline

Five phases. No magic.

Every page is grounded in scanned evidence, repaired into valid Markdown, and checked for source coverage before the site builds.

Scan
Plan
Generate
API Ref
Build
Select a phase to explore
Getting started

Three commands.

Initialize against any Python, JavaScript/TypeScript, Go, PHP, or Vue codebase. Set one API key. Generate. The chatbot inherits the same key automatically and runs embeddings locally via fastembed.

Read the full docs
bash
# 1 — install $ pip install deepdoc # 2 — initialize against your repo $ deepdoc init --with-chatbot --provider anthropic $ export ANTHROPIC_API_KEY=sk-ant-... # 3 — generate and preview $ deepdoc generate ✓ 47 pages generated in 1m 12s ✓ Chatbot index built (1,247 chunks) $ deepdoc serve Docs → http://localhost:3000 Chat → http://localhost:3000/ask

Your codebase already tells a story. Let it.

MIT licensed. No account. No cloud. Runs entirely in your terminal.