LLM Docs

Use AIOZ Stream documentation as context for AI coding assistants to generate better-grounded, SDK-aware code suggestions.

What is llms.txt?

llms.txt is a plain-text entry file that helps AI coding assistants discover and navigate AIOZ Stream documentation in a structured, machine-readable format. By providing this file as context, tools like Claude, Cursor, GitHub Copilot, and ChatGPT can identify the right documentation sections, follow relevant links, and generate better-grounded code using the appropriate SDK methods, parameters, and authentication patterns.

Downloads

Choose the file that fits your use case:

Index file: Links to all section files. LLMs can pick only the sections they need — best for context-length-sensitive tools.

curl -o aioz-stream-llms.txt https://aiozstream.network/docs/llms.txt

Full reference: Every endpoint and SDK example in one file. Best for one-shot context injection or tools with large context windows.

curl -o aioz-stream-llms-full.txt https://aiozstream.network/docs/llms-full.txt

Individual section files

Download only the sections you need. Each file contains SDK code examples for that API area.

How to use it

Claude (claude.ai or Claude Code)

I've attached the AIOZ Stream API reference (Node.js SDK).
Help me write a function that uploads a video and returns the HLS playback URL.

Cursor

@media-video.md

Create a TypeScript class that wraps video upload with progress reporting
and automatic encoding status polling.

ChatGPT

Step 1: Download the file above.

Step 2: Open chatgpt.com (opens in a new tab), start a conversation, and attach the file.

Step 3: Ask:

I've attached the AIOZ Stream developer documentation docs.
Generate a complete Node.js script that:
1. Creates a video object
2. Uploads a local file
3. Polls until encoding is done
4. Returns the HLS and DASH URLs

GitHub Copilot / VS Code

Step 1: Download the section file(s) relevant to what you're building.

Step 2: Open the file in your editor so it's part of the workspace context.

Step 3: Reference it in Copilot Chat using #:

#media-video.md

Write a TypeScript function that creates a live stream with 720p quality
and returns the RTMP ingest URL and HLS playback URL.

Example prompts

Copy these prompts after loading the docs as context:

Using the AIOZ Stream developer documentation docs, write a Node.js Express endpoint that:
- Accepts a video file upload via multipart/form-data
- Creates an AIOZ Stream video object
- Uploads the file using the SDK
- Returns the video ID and a polling endpoint for encoding status

Using the AIOZ Stream developer documentation docs, create a TypeScript webhook handler that:
- Receives encoding_finished events
- Extracts the video ID
- Fetches the video detail to get playback URLs
- Updates a database record (pseudo-code for the DB part is fine)

Using the AIOZ Stream developer documentation docs, write a React hook that:
- Takes a videoId prop
- Uses the SDK to fetch video detail
- Returns { hlsUrl, status, isLoading, error }

Using the AIOZ Stream developer documentation docs, implement a CLI script that:
- Accepts a folder path as argument
- Uploads all .mp4 files in the folder to AIOZ Stream
- Shows a progress bar for each upload
- Prints a summary with video IDs and HLS URLs when done

What's in the files

All files are plain text, without HTML or MDX, making them easier for AI tools to ingest and process.

llms.txt — index file following the llmstxt.org (opens in a new tab) standard:

• Behavioral instructions for AI assistants
• Links to each section file
• Link to the full combined reference

llms-full.txt — complete reference in one file:

• All sections concatenated
• Same SDK examples as individual files

Section files (llms/*.md) — one per API area:

• SDK method signatures with TypeScript
• Full end-to-end usage examples (create → upload → poll → use)
• Fallback fetch patterns for APIs not yet in the SDK
• Endpoint path, method, parameters, and response shape