dsh-model-retry-settings
Run the following command in DeepSeek Harness:
dsh plugin install y2zyyr/dsh-model-retry-settings
Paste the following prompt into your AI chat to install this plugin:
Install via DSH CLI by running dsh plugin install y2zyyr/dsh-model-retry-settings, with the full source available at https://github.com/y2zyyr/dsh-model-retry-settings; restart DSH Desktop after installation to load the plugin.
About this plugin
DSH Desktop ships with automatic model-request retries capped at a fixed limit of 2. On a flaky connection you may want more attempts; while debugging you may prefer instant failure over silent retries. The dsh-model-retry-settings plugin exposes that single knob in Settings → General, letting you pick from 0, 1, 2, 3, 5, or 10. The default remains 2, so installing the plugin without touching the setting is a no-op.
Use Cases
- Flaky network causes frequent model-request failures and you need a higher retry cap for reliability
- Debugging prompt or tool-calling pipelines with auto-retry disabled to see the raw error immediately
- Hitting rate-limit or 5xx bursts and wanting extra retry chances before a human intervenes
Best For
- Daily DSH Desktop users who frequently hit network jitter or transient server errors
- Platform developers debugging inference pipelines who need the raw error right away
- Team admins who prefer to deploy and switch model behavior via a visual settings UI
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.