trio-stream-vision

# Trio Stream Vision — Analyze Any Livestream with Natural Language

Paste a YouTube Live URL, RTSP camera feed, or HLS stream — ask questions about what's happening in plain English. Detect events, monitor conditions, and get periodic summaries, all without writing ML pipelines. Powered by [Trio's Reality-as-an-API](https://machinefi.com).

## When to Use

- User asks what's happening on a camera, stream, or video feed ("is anyone at my front door?")
- User wants smart alerts for specific events ("tell me when a package is delivered", "alert me if my dog gets on the couch")
- User wants to monitor something they can't watch themselves (construction site, parking spot, warehouse)
- User wants periodic summaries of a live feed ("summarize this stream every 10 minutes")
- User provides any live stream URL: YouTube Live, Twitch, RTSP/RTSPS cameras, HLS streams

## Prerequisites

- A Trio API key. Get one free (100 credits) at https://console.machinefi.com
- Set the key: `export TRIO_API_KEY=your_key_here`
- Base URL: `https://trio.machinefi.com/api`

## Available Actions

### 1. Check Once (Quick Snapshot)

Ask a yes/no question about what's currently visible on a stream. Costs 1 credit ($0.01).

```bash
curl -s -X POST "https://trio.machinefi.com/api/check-once" \
  -H "Authorization: Bearer $TRIO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "stream_url": "STREAM_URL_HERE",
    "condition": "NATURAL_LANGUAGE_CONDITION_HERE"
  }' | python3 -m json.tool
```

**Optional parameters:**
- `"include_frame": true` — returns the analyzed frame as base64 image
- `"input_mode": "clip"` — analyze a short video clip instead of a single frame (better for motion detection)
- `"clip_duration_seconds": 5` — clip length (1-10 seconds, only with clip/hybrid mode)

**Response fields:**
- `triggered` (boolean) — whether the condition matched
- `explanation` (string) — VLM's reasoning about what it sees
- `latency_ms` — processing time in milliseconds

**Input mode guidan...

# trio-stream-vision — Analyze Any Livestream with Natural Language

Paste a YouTube Live URL, RTSP camera feed, or HLS stream — ask questions about what's happening in plain English. Detect specific events, monitor conditions continuously, or get periodic summaries. No ML pipelines, no model downloads, no GPU required.

Powered by [Trio's Reality-as-an-API](https://machinefi.com) vision models.

## The Problem

Live video is everywhere — YouTube streams, security cameras, webcams, Twitch channels — but extracting useful information from them requires building ML pipelines, managing frame extraction, and serving vision models. Enterprise video AI starts at $50K+/year.

## The Solution

Describe what you want to know in plain English. Trio's vision models analyze the stream and return structured answers.

| What You Say | What Happens | Cost |
|-------------|-------------|------|
| "Is anyone visible on this stream?" | Instant yes/no + explanation of what the VLM sees | $0.01 |
| "Alert me when a delivery truck appears" | Continuous monitoring, triggers when condition is met | $0.02/min |
| "Summarize this livestream every 10 minutes" | Periodic narrative digests of stream activity | $0.02/min |

Works with **YouTube Live**, **RTSP/RTSPS cameras**, and **HLS streams**.

## Install

```bash
# From ClawHub
openclaw skills install machinefi/trio-stream-vision

# Or clone directly
git clone https://github.com/machinefi/trio-openclaw-skill.git ~/.openclaw/skills/trio-stream-vision
```

## Setup

1. Get a free API key (100 credits / $1.00) at https://console.machinefi.com
2. Set your key:
   ```bash
   export TRIO_API_KEY="your_key_here"
   ```
3. That's it. No ML setup, no model downloads, no GPU required.

## What You Can Do

**Instant Stream Analysis**
> "What's happening on this YouTube livestream right now?"
> "Is there anyone visible in the frame?"
> "How many cars are in the parking lot?"

**Continuous Event Detection**
> "Watch this stream and alert me when a perso...