Local-first, durable workflow runtime for Agent Apps—compile static Workflow DSL to LangGraph and orchestrate trusted multi-agent execution through Codex, WorkBuddy, DeepSeek Harness, or any MCP client.
A tiny glass companion for DeepSeek Harness — it breathes while the agent thinks, pulses while it works, and politely checks with you before touching anything. 🍎 Liquid Glass looks · 🌱 zero extra dependencies · 👀 always knows what your agent is up to;给 DeepSeek Harness 的一只小玻璃伙伴——Agent 思考时它轻轻呼吸,干活时它微微脉动,要动你的东西之前,它还会先礼貌地问你一声。
Busy-time input queue for the DeepSeek Harness web GUI: three-tier (now/next/later) planning, queue editing with drag-to-reorder and concurrency protection, interject (steer) and interrupt, batch clear with confirm, session freeze. A cordis client plugin, no dsh source changes, no PR required.
Per-agent model switching for DeepSeek Harness: set_model / list_models agent tools with provider whitelist and token-budget guardrails, plus plan-mode model staging and cache-safe active-model context.
DeepSeek Harness (DSH) plugin: a "Thinking Language" selector under Settings → General that tells the model which language to reason in (chain of thought) via a system-prompt section. Zero extra calls, zero latency, 12 languages.