Extra model-request recovery for the DeepSeek Harness: retries a failed model request once after a fixed delay when the provider classifies the failure as PI_AI_ERROR (unrecognized upstream error text). All other error codes stay on the built-in dsh-llm-retry default policy (5 bounded exponential-backoff retries). Composes safely with @deepseek-ai/dsh-llm-retry via the agent/request-error waterfall using its own policyKey bookkeeping.
Agent 협업
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:lim12137/dsh-llm-extra-retry
Provider failover for DeepSeek Harness: on rate-limit (429) or quota exhaustion, automatically switch between providers with configurable per-provider models, fall back to the last one (e.g. DeepSeek official), and show a visible notice in the UI.
인터페이스 확장
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:HB00/dsh-llm-failover
Provider/model fallback chain plugin for DeepSeek Harness — when the current model fails with an eligible error code (e.g. RATE_LIMIT or QUOTA), the agent loop automatically retries with the next provider/model in the configured chain.
Agent 협업 · 개발 도구
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:morphlinglan/dsh-llm-fallback
DSH harness plugin: tolerate OpenAI-compatible providers whose streaming responses omit finish_reason (e.g. the Snowflake Cortex gateway) so turns complete instead of failing with a TRANSPORT error.
Kilo Gateway free-model provider for DeepSeek Harness: auto-discovers the free model catalog, exposes tools/reasoning/image capabilities, records token usage, and gives you a settings page for your own API key and a catalog change log.
시각 도구
$ npx -p @deepseek-ai/dsh dsh plugin --profile web add github:aqiu817/dsh-llm-kilo-gateway
DeepSeek Harness plugin: serve LLM calls with the OAuth tokens your local Codex CLI and Claude Code already hold, instead of a separately configured API key.