dsh-llm-vision-bridge

Plugin overview

Native LLM-provider vision bridge: images pasted in the chat are described by a vision model (Qwen3-VL via pi-ai/llama.cpp) and the text description is fed to text-only DeepSeek for the reply — image admission, routing and compaction all run through harness-native mechanisms, with an LRU description cache and 503 retry.

Suitable scenarios

This guidance is based on the entry category, labels, and public information. Actual capabilities depend on installation and use in the GetRun Harness App.

• When the workspace needs to work with images, screenshots, or mixed image-and-text tasks, review this entry's description and labels such as Image understanding, Multimodal tasks to judge whether it fits the task.
• Validate whether "dsh-llm-vision-bridge" fits the current task in a test session without sensitive data before using it in an active project.
• When comparing entries in the same category, review public stars, version, license, and labels together. Those fields are not promises of security, compatibility, or ongoing maintenance.