dsh-tavern-renderer
Run the following command in DeepSeek Harness:
dsh plugin install LiweiDonVee/dsh-tavern-renderer
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install LiweiDonVee/dsh-tavern-renderer in your terminal to install the plugin; the source is available at https://github.com/LiweiDonVee/dsh-tavern-renderer .
About this plugin
DeepSeek Harness message bubbles are plain by default: rich Markdown ends up as flat black-on-white text with no visual layering. dsh-tavern-renderer plugs into the native Cordis keyed slot on the display side only, injecting an ordered rendering pipeline into assistant-step, user, and steering nodes without touching host source code. The pipeline covers regex replacement, scoped macro variables, a constrained inline expression evaluator (eval, function calls, prototype access, and constructors are all rejected), Markdown parsing, HTML/CSS whitelist sanitization, and semantic document templates. Every stage can be toggled off or flanked by custom extensions, and none of them modify persisted session events or the model prompt.
The standout feature is eight immersive document templates: a letter with fold crease and wax seal, an exam sheet with scoring boxes, a postcard with stamp and address lines, a classified cipher with redaction bars, a telegram with operator and priority fields, a newspaper clipping with edition and headline, a dossier with agency and handler, and a diary with weather and mood. A model only needs a short fenced block with a template name on the first line followed by regular Markdown; the plugin assembles the full visual structure. All metadata is text-escaped before rendering, and field values can never produce class names, styles, or attributes.
It suits users running role-play, long-form narratives, or document-style outputs in Harness who want SillyTavern-grade visual richness without introducing arbitrary JavaScript execution.
Use Cases
- Rendering immersive letters, exam sheets, and telegrams in Harness chat
- Filling dynamic message content with macros and safe inline expressions
- Customizing themes and CSS without affecting the host application
Best For
- Developers running role-play or long-form narratives in Harness
- Users wanting visual richness in message display without arbitrary JS
- Integrators needing safe rendering of third-party Markdown and HTML
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.