A dedicated vision skill that brings Kimi K2.5 image recognition capabilities to Alibaba DashScope GLM users within the OpenClaw ecosystem.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install openclaw-dashscope-kimi2-5-vision
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install openclaw-dashscope-kimi2-5-vision using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Kimi K2.5 Vision Integration?
This skill bridges the technical gap for users of Alibaba DashScope GLM models by providing high-performance image understanding through Kimi K2.5. It is specifically designed to handle scenarios where the primary model lacks native media processing capabilities, ensuring your AI agent can interpret visual data seamlessly.
By integrating this into your suite of Openclaw Skills, you enable robust vision-based workflows without manual configuration overhead. The skill intelligently manages authentication by searching through multiple potential configuration sources, making it a plug-and-play solution for developers building sophisticated multi-modal agents.
Kimi K2.5 Vision Integration Use Cases
- Identifying objects or text within images when the primary LLM returns a no media-understanding provider error.
- Automating OCR tasks to extract text from screenshots, UI designs, or technical documents.
- Enhancing coding agent workflows with visual context from diagrams or architectural charts.
- Supporting multi-modal interactions in custom AI agent implementations within the OpenClaw ecosystem.
How Kimi K2.5 Vision Integration Works
- The skill monitors for specific triggers such as the no media-understanding provider error or direct user image analysis requests.
- It initiates an automated API key discovery sequence, checking environment variables and multiple OpenClaw configuration paths.
- Once the DashScope API key is identified, it caches the credentials to the local environment for future sessions.
- The recognize.py script processes the provided image path and prompt using the Kimi K2.5 vision model.
- The resulting analysis or text extraction is returned to the agent or user to proceed with the workflow.
Kimi K2.5 Vision Integration Setup
First, ensure you have your Alibaba DashScope API key from the Bailian console. You can then configure it by adding it to your environment:
echo 'DASHSCOPE_API_KEY=sk-sp-your_key_here' >> ~/.openclaw/.env
To run a manual recognition test, use the following command:
python3 scripts/recognize.py /path/to/image.jpg --prompt "Describe this image"
Kimi K2.5 Vision Integration Data Schema & Taxonomy
The skill interacts with the following configuration and environment data to maintain state and authentication:
| Component | Description |
|---|---|
| DASHSCOPE_API_KEY | Primary environment variable for authentication. |
| ~/.openclaw/.env | Persistent storage for API credentials and local configuration. |
| recognize.py | The core execution script for processing images via the Kimi K2.5 model. |
| Input Parameters | Supports local file paths and custom string prompts for targeted analysis. |
Kimi K2.5 Vision Integration Advanced Features
- Automated credential fallback across multiple configuration paths including environment variables and OpenClaw provider blocks (qwencode or dashscope).
- Intelligent trigger detection to automatically fill gaps in LLM media capabilities when a model fails to understand visual data.
- Persistent credential caching to minimize setup friction across different Openclaw Skills environments.
- Support for custom prompts to tailor the vision output for specific technical, data extraction, or creative tasks.
SKILL.md
Loading
Related Openclaw Skills
Vaporwave Art Generator
An AI-powered creative tool for generating retro-aesthetic, synthwave, and vaporwave images using simple text prompts.
Short-Video-AI Pipeline
An automated production pipeline that transforms a single topic or idea into a fully edited 60-90 second vertical video delivered via Telegram.
Claw Relay CLI Client
A stateless CLI client that allows AI agents to control real, authenticated remote browsers through simple execution commands.
Tech Solution Research
A professional technical research skill that orchestrates multi-source evidence for framework selection and technology comparison.
Poker AI Agent Skill
A high-performance gaming skill that enables AI agents to play No-Limit Texas Hold'em against other agents in a real-time casino environment.
MLM Assistant
A comprehensive automation tool for network marketing professionals to manage prospecting, follow-ups, team onboarding, and FTC compliance.
Featured*
