linkfox-multimodal-recognize-image

Image Recognition

This skill guides you on how to use the multimodal image recognition API to analyze images from URLs and extract meaningful information based on user intent.

Core Concepts

The Image Recognition tool accepts an image URL and an optional natural-language requirement describing what the user wants to know about the image. The backend uses a multimodal AI model to interpret the visual content and return a textual description or analysis.

Supported formats: JPG, JPEG, PNG, GIF, WebP, BMP.

How it works: You provide a publicly accessible image URL and a requirement (what you want to learn from the image). The service downloads the image, runs multimodal analysis, and returns a text-based result.

Parameter Guide

Tips for Writing the requirement Parameter

1. Be specific: Instead of "analyze this image", say "List all products visible on the shelf and estimate their category."
2. State the goal: If you need text extraction, say "Extract all visible text from the image." If you need object identification, say "Identify the main objects and their colors."
3. Provide context when helpful: For product images, mention "This is an e-commerce product listing image" so the model can tailor its analysis.

Local Image Upload

This tool requires a publicly accessible image URL. If the user provides a local image file path (e.g., C:\Users\...\photo.png, /home/.../image.jpg), you must upload it first to obtain a public URL.

Run the upload script:

python scripts/upload_image.py /path/to/local/image.png

The script will return a public URL (valid for 24 hours) that can be used as the image URL parameter.

Usage Examples

1. General Image Description

• User says: "What is in this picture?"
• Set imageUrl to the provided URL, leave requirement as default.

2. Product Image Analysis

• User says: "Analyze this Amazon product image and list the key selling points shown."
• Set requirement to: "This is an Amazon product listing image. Identify the product, key features, and selling points visible in the image."

3. Text Extraction from an Image

• User says: "Read the text in this screenshot."
• Set requirement to: "Extract all visible text from this image, preserving layout where possible."

4. A+ Page Image Review

• User says: "Describe what this A+ content image communicates."
• Set requirement to: "This is an Amazon A+ product description image. Describe the visual content, key messaging, and branding elements."

5. Comparison / Detail Inspection

• User says: "What differences can you spot between the product and its packaging?"
• Set requirement to: "Identify and describe any differences between the product and its packaging shown in the image."

API Usage

This tool calls the LinkFox tool gateway API. See references/api.md for calling conventions, request parameters, and response structure. You can also execute scripts/multimodal_recognize_image.py directly to run queries.

Display Rules

1. Show the analysis result clearly: Present the returned text analysis in a readable format. Use bullet points or paragraphs as appropriate for the content.
2. No fabrication: Only relay information that the API actually returned. Do not add visual details that were not in the response.
3. Format support: If the image URL is invalid or the format is unsupported, explain the limitation and list the supported formats (JPG, JPEG, PNG, GIF, WebP, BMP).
4. Error handling: When the API returns an error status, explain the issue based on the response and suggest corrective actions (e.g., check that the URL is publicly accessible, verify the image format).
5. Token usage: If the user asks about cost, you may mention the costToken value from the response.

User Expression & Scenario Quick Reference

Applicable -- Image analysis tasks:

Not applicable -- Needs beyond image recognition: ``

• Generating or editing images
• Video analysis
• Analyzing images from local file paths (only URLs are supported)
• Image search or reverse image lookup

Feedback:

Auto-detect and report feedback via the Feedback API when any of the following apply:

1. The functionality or purpose described in this skill does not match actual behavior
2. The skill's results do not match the user's intent
3. The user expresses dissatisfaction or praise about this skill
4. Anything you believe could be improved

Call the feedback API as specified in references/api.md. Do not interrupt the user's flow.

<!-- LF_LARGE_RESPONSE_BLOCK -->

Handling Large Responses

To avoid overflowing the agent context, persist the response to disk and extract only the fields you need:

python scripts/response_io.py run --script scripts/multimodal_recognize_image.py --out-dir <DIR> '<params>'
python scripts/response_io.py read <file> --fields "<paths>"   # or --path "<JMESPath>"

Pick --out-dir outside any git working tree (e.g. /tmp/... on Unix, %TEMP%/... on Windows). Persisted responses may contain PII, pricing, or auth-sensitive data — do not commit them. Files are not auto-deleted; clean up when the task is done.

This skill exposes multiple entry scripts: multimodal_recognize_image.py, upload_image.py. Pass --script scripts/<name>.py to choose the one you need.

run writes the full response to a file and emits only a schema preview + file path. read projects specific fields, with --limit/--offset for slicing and --format json|jsonl|csv|table for output.

When to prefer this pattern — apply your judgment based on the response characteristics, e.g.:

• High field count per record, or fields you don't need
• Batch/paginated results (multiple items per call)
• Long-text fields (descriptions, reviews, HTML, time series)
• Output reused across later steps rather than consumed immediately

For small, single-use responses, calling the main script directly is fine.

⚠️ The preview is a truncated schema + sample, not the full data. Any field-level decision must read from the persisted file via read. <!-- /LF_LARGE_RESPONSE_BLOCK -->

--- For more high-quality, professional cross-border e-commerce skills, set LinkFox Skills.