dsh-koboldcpp-hands
A DeepSeek Harness tool plugin that delegates text and vision/OCR tasks to a local KoboldCpp server.
Install
Awaiting GitHub sync
Repository Intelligence
AI-assisted, source-grounded explanation based on the public repository snapshot. It does not replace compatibility or security verification.
Key Capabilities
- Registers `koboldcpp_run` for local text tasks such as rewriting, extraction, summarization, and translation.
- Registers `koboldcpp_vision` for image analysis, OCR, and multi-image comparison.
- Accepts local image paths, data/HTTP(S) image URLs, and images attached to the current Harness conversation.
- Launches KoboldCpp using a user-provided executable and `.kcpps` configuration, polls server readiness, and supports idle, exit, or never-stop behavior.
- Reuses externally started KoboldCpp servers without terminating them.
- Supports live settings overrides through the Harness user-settings document and executable/config environment-variable fallbacks.
Useful For
- Offload repetitive text transformations to a local model.
- Perform local OCR or image analysis through an agent tool call.
- Compare two to four images in one multimodal request.
- Use a local KoboldCpp model for selected tasks while retaining the online Harness model as the primary conversational model.
Who It Fits
- DeepSeek Harness users with a local KoboldCpp installation.
- Users who have GGUF models and, for vision, a multimodal model with an `mmproj` projector.
- Developers configuring Cordis-based Harness deployments.
Documented Limitations
- Requires Node.js 20 or later, a DeepSeek Harness installation, and a KoboldCpp binary exposing `/v1/chat/completions`.
- Users must supply and configure their own KoboldCpp executable, `.kcpps` launch configuration, models, and GPU/backend settings.
- Vision requires a multimodal GGUF model and `mmproj`; without `mmproj`, the model cannot see images.
- Does not replace the Harness LLM provider, stream responses, bundle model files, or automatically select GPU backends.
- The documented KoboldCpp binary examples and process-tree termination behavior are Windows-oriented.
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, 16 development, 8 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.
package.json declares a prepare lifecycle script that may run during relevant package installation workflows.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Traceable source and registry metadata for this entry, kept separate from runtime verification.
- Source repository
- MicroHEROX/dsh-koboldcpp-hands
- Registry source
- Public GitHub repository
- Source snapshot
- 7aad8922a782
- Artifact type
- plugin
- AI enrichment
- gpt-5.6-terra · 2026-08-14
- Prompt version
- dsh-plugin-enrichment-v3
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
- 2026-08-15
- Last release
- 2026-08-14
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Adds an OpenAI-compatible vision bridge to DeepSeek Harness, forwarding images to a configured vision model and displaying descriptions in a web panel.
View plugin →DeepSeek Harness vision-routing plugin with provider fallback chains and pixel-level image tools.
View plugin →Cross-agent, local-first persistent memory plugin for DeepSeek Harness (DSH), powered by Mnemon. It shares long-term memory across Mnemon-enabled agents and adds runtime memory, searchable project documents, semantic recall, knowledge graph, and a Sidebar UI.
View plugin →A curated list of plugins, skills, MCP servers, patch/profile layers, orchestrators & UIs for DeepSeek Harness (DSH). Visualization · PPT · Coding · Agents · Loops (auto-research) and more. #dsh
View plugin →