Fixes missing reasoning_content for DeepSeek V4
⚡️ 零 API 账单！基于 OpenAI Prism 的无限 GPT Token 反代服务，直通满血 o1-astra-xhigh / o3 深度思考链，全兼容 OpenAI 接口。
Restores reasoning content dropped from multi-turn OpenCode conversations (DeepSeek, Kimi, GLM, MiMo, MiniMax, InclusionAI/Ling, OpenRouter). V1 + V2 plugin, proxy, and watchdog.
(WORKS) NSFW/Coding refusals removing local proxy. If ENI is not enough.
Fix missing chain-of-thought in WorkBuddy (and any client reading reasoning_content) when using vLLM >= 0.28, which renamed the field to reasoning
Self-built ReAct loop for DeepSeek. First to round-trip reasoning_content as core design (40-60% token savings). Zero-framework TS+Bun binary. Part of the Loop Engineering family.
Compatibility and diagnostics for DeepSeek V4 tool-calling agents
面向 Claude Code 和 Anthropic 兼容客户端的小米 MiMo 本地代理：修复 reasoning_content 400，兼容 1M 别名/count_tokens，支持 macOS/Windows 一键启动。Local Anthropic-compatible Xiaomi MiMo proxy for Claude Code and coding agents: fixes reasoning_content 400 errors, handles 1M aliases/count_tokens, and ships macOS/Windows builds.
Open reasoning for health-literacy tools — plain-language explanations of a person's own health data, structured for expert review.
To associate your repository with the reasoning-content topic, visit your repo's landing page and select "manage topics."