DeepDocs: Fix Your Outdated Docs On Autopilot 🔥

DeepDocs is an AI agent that automatically keeps your docs like READMEs, API references, SDK guides, examples & tutorials in-sync with your codebase, eliminating manual effort.

🛠️ How it works (Watch)

DeepDocs works by creating a rich mapping between your code and docs. When you commit changes, it automatically detects and updates the relevant doc files.

✨ Features

• 🚀 Continuous Documentation: Automatically scans for outdated docs after every commit and proposes updates.
• 🔍 Deep Scan: Entire repo is scanned on every commit, not just a few files
• 🧠 Intelligent Updates: Makes precise edits without changing your formatting, tone, or structure.
• 📊 Detailed Reporting: Explains what changed, why, and which commits triggered each update.
• 🧩 Works With All Setups: Supports both monorepos, and separate doc repos. Blends easily with doc generator tools.
• 🔄 GitHub Native: Fits directly into your GitHub development workflow

⚡ Quickstart

1️⃣ Install DeepDocs on your source repo (and docs repo if separate)

2️⃣ Add a deepdocs.yml to the root of your repo, in the branch you're working (e.g. main or a feature branch)

target:  
  - url: "docs/"
  - url: "examples/"
include: []
reinit: "random-string"

where

• target.url: Relative path or GitHub URL to the docs file/folder
• include: Optional list of files or folders to monitor. Leave empty to monitor the whole repo.
• ignore: Optional list of relative paths to skip

3️⃣ Commit it

What happens next:

• DeepDocs initializes and runs a deep scan to detect outdated docs.
• Gives you a report on which docs are outdated and why
• Creates a branch: deepdocs-update-<branch>-<commit_sha> with proposed updates. Here, <branch> is your current branch name and <commit_sha> is the commit of deepdocs.yml

Every subsequent commit will trigger an automatic scan and fix of outdated docs as described above.

📈 Credits Usage

Credits are consumed based on the number of doc files processed by the LLM during a scan, and can vary between 0 and the total number of doc files that require analysis.

🔗 Important Links

• 📘 Docs
• 🌐 Website
• 📧 Support: neel.das@deepdocs.dev