---
name: camscanner-image-detect-aigc
description: Use CamScanner to detect whether an image was generated by an AI model (e.g. Stable Diffusion, Midjourney, DALL·E). Powered by an AIGC-detection engine that classifies an image as genuine, suspected AI-generated, or AI-generated, with a confidence score. Returns a JSON result containing `ai_check_result` (1/2/3), `confidence`, and `result_text`. Use when the user asks whether a photo is AI-generated, wants to verify an image's authenticity against AI generation, or asks "is this AI art / Stable Diffusion / Midjourney?". Triggers on "检测AI生成", "是不是AI画的", "AIGC检测", "AI图片识别", "detect AI-generated image", "is this AI art", "is this diffusion / midjourney", or when the user shares an image and asks whether it was produced by AI.
metadata:
  author: CamScanner
  version: "1.0"
  openclaw:
    emoji: "🤖"
    requires:
      bins: ["curl", "jq"]
  homepage: "https://www.camscanner.com"
---

# CamScanner Image Detect AIGC

## Overview

CamScanner provides an AIGC-detection engine that determines whether an image was produced by an AI generator. The workflow is a 2-step pipeline: **upload** the image, then **validate** it with `validate_mode: 2`. Unlike conversion skills, this skill does not produce a file — the validate step returns a JSON result whose key fields (`ai_check_result`, `confidence`, `result_text`) should be reported back to the user directly.

## When to Use

- User asks whether an image was generated by AI (Stable Diffusion, Midjourney, DALL·E, etc.)
- User wants to distinguish a real photo from AI-generated artwork
- User asks "is this AI art / AI-generated?" or similar authenticity questions
- User shares an image and explicitly asks whether it came from a generative model

## Presenting the Result

- Always read `ai_check_result`, `confidence`, and `result_text` from the response and report them in plain language.
- Map `ai_check_result` to a verdict: `1` = not AI-generated, `2` = suspected AI-generated, `3` = AI-generated.
- Include the `confidence` value (a float in `[0, 1]`) so the user understands how certain the verdict is.
- **Match the user's language.** `result_text` is returned in Chinese by the API. If the user asked in English (or any other language), translate/rephrase it into that language. If the user asked in Chinese, you can use `result_text` as-is.
- Do not overstate the verdict: "suspected AI-generated" is not the same as "AI-generated".

## Privacy & Data

> **Important: Privacy & Data Flow Notice**
>
> - **Third-party service**: This skill sends your files to CamScanner's official servers (`ai-tools.camscanner.com`) for processing.
> - **Data retention**: CamScanner servers process your files in real-time. Files are not permanently stored on the server.
> - **Result**: Only a JSON detection result is returned — no file is downloaded.

## API Reference

**Base URL:** `https://ai-tools.camscanner.com`

### Supported Validations

| source_type | validate_mode | Detection               | Engine          |
| ----------- | ------------- | ----------------------- | --------------- |
| image       | 2             | AIGC (AI-generated)     | aigcdetection   |

### Step 1: Upload Image

```bash
BASE="https://ai-tools.camscanner.com"

IN_FILE_ID=$(curl -sS -X POST "$BASE/v1/tools/upload_file/execute" \
  -H "Content-Type: application/octet-stream" \
  --data-binary "@/path/to/image.jpg" | jq -r '.tool_result.data.file_id')
```

**Response:**

```json
{
  "code": 200,
  "tool": "upload_file",
  "tool_result": {
    "success": true,
    "data": {
      "file_id": "file_1741857600_ab12cd34ef56.jpg",
      "size": 24576
    }
  }
}
```

### Step 2: Validate Image (Detect AIGC)

```bash
curl -sS -X POST "$BASE/v1/tools/validate_image/execute" \
  -H "Content-Type: application/json" \
  -d "{\"file_id\":\"$IN_FILE_ID\",\"validate_mode\":2}"
```

**Response (suspected AI-generated example):**

```json
{
  "code": 200,
  "tool": "validate_image",
  "tool_result": {
    "success": true,
    "data": {
      "ai_check_result": 2,
      "confidence": 0.346435546875,
      "engine": "aigcdetection",
      "file_id": "file_xxx.jpg",
      "result_text": "检测结果为疑似 AI 生成图片",
      "review_state": "auto_checked",
      "validate_mode": 2
    },
    "metadata": {
      "ai_check_result": 2,
      "confidence": 0.346435546875,
      "engine": "aigcdetection",
      "result_text": "检测结果为疑似 AI 生成图片",
      "review_state": "auto_checked",
      "validate_mode": 2
    }
  }
}
```

## Interpreting the Result

| Field             | Type    | Meaning                                                                        |
| ----------------- | ------- | ------------------------------------------------------------------------------ |
| `ai_check_result` | integer | `1` = not AI-generated, `2` = suspected AI-generated, `3` = AI-generated       |
| `confidence`      | float   | Model confidence score in `[0, 1]`                                             |
| `result_text`     | string  | Human-readable conclusion (Chinese by default — translate for other languages) |
| `review_state`    | string  | Review status (e.g. `auto_checked`) — informational, not user-facing           |
| `validate_mode`   | integer | Echo of the requested mode (always `2` for AIGC detection)                     |

### Verdict Mapping

| `ai_check_result` | Verdict                 | Suggested phrasing (EN)              |
| ----------------- | ----------------------- | ------------------------------------ |
| 1                 | Not AI-generated        | "Looks like a real image"            |
| 2                 | Suspected AI-generated  | "Suspected to be AI-generated"       |
| 3                 | AI-generated            | "Detected as AI-generated"           |

## Quick Reference: Complete Pipeline

Detect whether an image is AI-generated (two-step, reads JSON result):

```bash
BASE="https://ai-tools.camscanner.com"
INPUT_IMAGE="/path/to/image.jpg"

# Upload
IN_FILE_ID=$(curl -sS -X POST "$BASE/v1/tools/upload_file/execute" \
  -H "Content-Type: application/octet-stream" \
  --data-binary "@$INPUT_IMAGE" | jq -r '.tool_result.data.file_id')

# Validate and extract key fields
RESULT=$(curl -sS -X POST "$BASE/v1/tools/validate_image/execute" \
  -H "Content-Type: application/json" \
  -d "{\"file_id\":\"$IN_FILE_ID\",\"validate_mode\":2}")

AI_CHECK=$(echo "$RESULT" | jq -r '.tool_result.data.ai_check_result')
CONFIDENCE=$(echo "$RESULT" | jq -r '.tool_result.data.confidence')
RESULT_TEXT=$(echo "$RESULT" | jq -r '.tool_result.data.result_text')

echo "ai_check_result: $AI_CHECK"
echo "confidence:      $CONFIDENCE"
echo "result_text:     $RESULT_TEXT"
```

## Common Mistakes

| Mistake                                    | Fix                                                                         |
| ------------------------------------------ | --------------------------------------------------------------------------- |
| Wrong Content-Type on upload               | Upload uses `application/octet-stream`, not `multipart/form-data`           |
| Using GET instead of POST                  | Both endpoints use POST                                                     |
| Passing `validate_mode` as a string        | `validate_mode` is an integer — use `2`, not `"2"`                          |
| Including `output_mode` in the request     | `validate_image` does not use `output_mode`; it always returns JSON         |
| Treating `ai_check_result` as boolean      | It is a 3-state integer (`1`/`2`/`3`); map explicitly to a verdict          |
| Reporting "suspected" as "AI-generated"    | `ai_check_result == 2` means *suspected*, not confirmed — phrase accordingly |
| Reporting `result_text` verbatim in EN     | `result_text` is Chinese; translate to match the user's language            |

## Error Handling

Check each step before proceeding:

```bash
# After upload
if [ -z "$IN_FILE_ID" ] || [ "$IN_FILE_ID" = "null" ]; then
  echo "Upload failed"; exit 1
fi

# After validate
if [ "$AI_CHECK" = "null" ] || [ -z "$AI_CHECK" ]; then
  echo "Validation failed"; exit 1
fi
```