DeepSeek Harness Plugins
DeepSeek Harness PluginManifest Valid

dsh-vision-bridge

Universal vision bridge for DeepSeek Harness that routes and describes images for text-only LLMs.

VisionUI & ProductivityDeveloper ToolsBrowser & WebSecurity & Policy
1GitHub Stars0ForksUpdated2026-09-18

Install

$ dsh plugin --profile web add @goodandready/dsh-vision-bridge

Plugin Overview & Capabilities

AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.

source-grounded
dsh-vision-bridge is an installable plugin for DeepSeek Harness that bridges visual content to text-only language models. When non-vision chat models receive image inputs, the bridge processes images using configured vision backends (DSH catalog, OpenAI-compatible, Ollama, LM Studio, or custom webhooks) and provides either automatic text rewrites or 25 dedicated visual analysis tools.

Key Capabilities

  • Automatic image rewriting into text descriptions in hybrid or LLM modes
  • 25 visual tools covering grounding, cropping, object detection, image comparison, OCR, UI layout breakdown, and visual Q&A
  • Multi-channel routing across DSH catalog, OpenAI-compatible APIs, Ollama, LM Studio, and custom webhooks with fallback strategies
  • Local OCR fallback via Tesseract and optional HTML screenshot rendering via headless Chrome
  • LRU description caching and persistent evidence storage
  • Bundled vision-skills playbooks registered through the DSH skills provider
  • Web UI settings card for channel configuration, benchmarking, and testing

Useful For

  • Enabling text-only DeepSeek models to understand image attachments without turn failures
  • Extracting structured text, UI component hierarchies, and OCR from screenshots or documents
  • Performing local visual tasks via Ollama or LM Studio endpoints with automated failover

Who It Fits

  • DeepSeek Harness users running text-only LLMs who need multimodal vision handling
  • Developers needing programmatic visual inspection tools and OCR pipelines inside DSH

Documented Limitations

  • Requires external system binaries for select tools: Chrome for HTML screenshots, ffmpeg for video frames, pdftoppm for PDFs, and tesseract for local OCR (each degrades gracefully when missing)
  • Image size restricted by maxImageBytes configuration (default 20MB)

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

No automated security signals have been published yet.

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository
GooDAnDReaDY/dsh-vision-bridge
Registry source
GitHub · dsh-plugin topic
Registry classification
Plugin
Source checked
ba34ca1 · 2026-09-18

This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.

Repository Activity

GitHub stars
1GitHub stars
Forks
0
Open issues
0
Last commit
2026-09-18
Last release
2026-09-18
For maintainers

Maintaining this plugin?

This listing is generated from public repository data. If you maintain this project, you can review the information and share this listing with your users if you find it useful.

Add to README
Listed on DSHPlugin.app