Yaing-Yan/dsh-better-retry

README 摘要

dsh-better-retry 🔁 A DeepSeek Harness (DSH) Cordis plugin that retries any model-request failure — whatever the error code — with durable, backoff-scheduled retries, plus Settings → General sliders to tune the retry budget (0–64, default 8) and the 429 wait (5–120 s, default 15 s) live. 一个 DSH（DeepSeek Harness）Cordis 插件：无论报什么错都自动重试模型请求（持久化 + 指数退避），并在 设置 → 常规 里提供滑块实时调整重试次数（0–64，默认 8）与 429 等待时长（5–120 秒，默认 15 秒）。 ✨ Why / 为什么需要它 The stock dsh-llm-retry plugin only retries failures whose code appears in the provider route's retryPolicy.retryableCodes (default: EMPTY RESPONSE , RATE LIMIT , SERVER , TIMEOUT , TRANSPORT ). Everything else — AUTH , QUOTA , INVALID REQUEST , PI AI ERROR , LLM STREAM IDLE TIMEOUT , … — ended the turn immediately with "本轮运行失败". 内置的 dsh-llm-retry 只重试 provider 策略里列出的错误码（默认：空响应、限流、5xx、超时、传输错误）。其它错误（鉴权、配额、无效请求、流空闲超时……）一遇即终，直接「本轮运行失败」。本插件把这张白名单变成"全量重试"。 - Any error code — retries every failure class, capped by your budget. - Durable & visible — appends the same llm/retry / llm/retry-started events as the stock plugin, so the chat UI shows "正在重试模型请求（n/N）· Xs" and sessions stay replay-safe. - Sliders in Settings — Settings → General → "失败重试次数 / Failure retries" (0–…

分类