dsh-autovision
Adds automatic image transcription for text-only models in the DeepSeek Harness web UI using a configured multimodal model.
Install
$ dsh plugin --profile web add @iroam2375/dsh-autovisionPlugin 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.
Key Capabilities
- Registers image-capable twin adapters for pure-text models at runtime.
- Automatically redirects image-bearing agent requests to the appropriate twin adapter.
- Transcribes pasted image content through a configured multimodal model and injects the result as text.
- Provides the agent-callable autovision_read_image tool for reading image files with a task-specific instruction.
- Supports an optional provider whitelist and custom vision prompt.
- Uses an LRU cache for image transcriptions.
Useful For
- Paste screenshots or images into sessions using text-only DeepSeek or GLM models.
- Let an agent inspect an image file during a task.
- Keep a text-model workflow while delegating image recognition to an already configured multimodal model.
- Extract visible text or describe UI state from attached images.
Who It Fits
- DeepSeek Harness web users with text-only and multimodal models configured.
- Agents and developers who need image-to-text bridging without changing model configuration.
- Users who want to avoid built-in API keys or third-party relay services.
Documented Limitations
- Requires dsh web 0.1.0-rc.6 or later.
- Requires the user to configure a default multimodal vision model.
- A new session whose first message contains an image is silently skipped; sending a text message first is required.
- Images are limited to 5 MB by the attachment-local default.
- Transcription latency depends on the selected vision model.
- The settings page may display a cosmetic bare autovision provider row with no address.
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 0 development, 0 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- Junkrat9527/dsh-autovision
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- d6e9b13 · 2026-08-25
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
- 1
- Last commit
- 2026-08-18
- Last release
- No release detected
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Compatibility bridge and host ABI enabling unmodified Pi plugins to run natively on DeepSeek Harness.
View plugin →Local-first AI agent runtime and DSH plugin bundle providing sandboxed sessions, MCP tools, audit logs, and session replay.
View plugin →Preview, create, and edit spreadsheets, docs, slides, and canvases in DeepSeek Harness powered by Univer.
View plugin →Vision routing and pixel-level visual tool suite for DeepSeek Harness agents with built-in free fallbacks.
View plugin →