y2zyyr/dsh-model-retry-settings

y2zyyr★ 0JavaScript最后同步: 2026-08-19

在 GitHub 打开

Configure automatic model request retry limits directly from DSH Desktop settings.

README 摘要

dsh-model-retry-settings Configure the maximum number of automatic model-request retries directly from DSH Desktop. 中文说明 (README.zh-CN.md) Features - Native DSH Desktop Settings integration — a preference row in Settings → General . - Global retry limit — one value applies to every model provider that uses DSH's normal retry policy. - Values: 0 / 1 / 2 / 3 / 5 / 10 - Default: 2 (identical to DSH's built-in default — installing the plugin changes nothing until you configure it) - Persistent configuration — survives closing/reopening Settings and app restarts. - Live configuration updates — changes apply to the next failed model request without restarting. - Preserves DSH's existing behavior : retryable-error classification, exponential backoff / jitter / provider Retry-After handling, request cancellation, and conversation retry rendering. - No DSH core patch required — a pure plugin. Where the setting appears Chinese: Installing the plugin without changing the setting preserves DSH's existing default behavior: 2 . Retry semantics The configured value is the number of retries after the initial model request. It does not count the initial request. - 0 = initial request only (no autom…

在 GitHub 查看完整 README →
工具/开发deepseek-harnessdshdsh-plugin

分类