← Back to Skills Marketplace
SiliconFlow Qwen Vision
by
lixingfeng
· GitHub ↗
· v0.2.0
· MIT-0
455
Downloads
0
Stars
3
Active Installs
2
Versions
Install in OpenClaw
/install siliconflow-qwen-vision
Description
图片理解与分析。当用户需要分析图片内容、识别图片中的物体、描述图片场景、理解图片含义时使用此技能。支持图片问答、物体识别、场景描述等。
Usage Guidance
This skill is functionally consistent with an image-analysis tool, but exercise caution before installing: 1) The SKILL.md and script require an API key (OPENAI_API_KEY or --api-key) but the registry metadata does not declare this — confirm where you should obtain a SiliconFlow key and do not reuse your primary OpenAI or cloud API keys. 2) The script will POST your entire image (base64) to https://api.siliconflow.cn; only use non-sensitive images unless you trust that endpoint. 3) The package has no homepage or verified source — verify the publisher or prefer an official client. 4) Ensure Python and the 'requests' library are installed in a controlled environment. If you decide to proceed, create a dedicated, limited-scope API key for this use, test with harmless images first, and review network traffic/logs to confirm expected behavior.
Capability Analysis
Type: OpenClaw Skill
Name: siliconflow-qwen-vision
Version: 0.2.0
The skill is a straightforward implementation for image analysis using the SiliconFlow Qwen2.5-VL API. The script `scripts/analyze_image.py` correctly handles image encoding and communicates with the legitimate SiliconFlow endpoint (api.siliconflow.cn) using standard practices. No evidence of data exfiltration, malicious execution, or prompt injection was found.
Capability Assessment
Purpose & Capability
The name/description (image understanding using Qwen2.5-VL) aligns with the included script, which reads an image, base64-encodes it, and sends it to https://api.siliconflow.cn using a Qwen model. However the registry metadata declares no required credentials while the SKILL.md and script require an API key (OPENAI_API_KEY or --api-key). Also no homepage or authoritative source is provided.
Instruction Scope
Runtime instructions only direct the agent/user to read a local image file, optionally read an API key from OPENAI_API_KEY (or accept --api-key), and send the image and prompt to SiliconFlow's chat/completions endpoint. There is no evidence of unrelated file access, broad system probing, or data collection beyond the image and prompt.
Install Mechanism
No install spec (instruction-only plus a Python script) — this is low-risk in terms of installation. The script imports 'requests' but the package/dependency is not declared anywhere; users will need to ensure Python and the requests package are available. No remote archives or installers are fetched by the skill itself.
Credentials
The script and SKILL.md require an API key via OPENAI_API_KEY or --api-key, but the registry metadata lists no required env vars and no primary credential. This mismatch is a red flag: the skill will send your API key to an external domain (api.siliconflow.cn). The OPENAI_API_KEY name may encourage reusing an OpenAI key; users should avoid reusing high-privilege keys and prefer issuing a dedicated SiliconFlow key if available.
Persistence & Privilege
The skill does not request persistent presence (always: false) and does not modify other skills or system-wide config. It runs on demand and does not claim elevated platform privileges.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install siliconflow-qwen-vision - After installation, invoke the skill by name or use
/siliconflow-qwen-vision - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.2.0
修复错误的 Minimax 注释，更新为正确的 SiliconFlow Qwen2.5-VL 描述
v0.1.0
siliconflow-qwen-vision 0.1.0
- Initial release with image understanding and analysis capabilities using SiliconFlow Qwen2.5-VL.
- Supports object recognition, scene description, and image-based Q&A.
- Allows configuration via environment variable or command-line API Key.
- Provides example usage for image description, object identification, scene analysis, and result saving.
Metadata
Frequently Asked Questions
What is SiliconFlow Qwen Vision?
图片理解与分析。当用户需要分析图片内容、识别图片中的物体、描述图片场景、理解图片含义时使用此技能。支持图片问答、物体识别、场景描述等。 It is an AI Agent Skill for Claude Code / OpenClaw, with 455 downloads so far.
How do I install SiliconFlow Qwen Vision?
Run "/install siliconflow-qwen-vision" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is SiliconFlow Qwen Vision free?
Yes, SiliconFlow Qwen Vision is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does SiliconFlow Qwen Vision support?
SiliconFlow Qwen Vision is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created SiliconFlow Qwen Vision?
It is built and maintained by lixingfeng (@tomlee2013); the current version is v0.2.0.
More Skills