Modelli e provider

Keeps long DSH sessions alive when a model channel starts failing: daruma takes over after the in-box retry gives up, trips a self-healing circuit breaker on repeated 429/5xx or a terminal error, and moves the next request to another channel in your chain. A panel beside the composer picks the backup by hand, switches show up as rows in the conversation, and every decision lands in a durable JSONL audit log.

Three independent language channels for model output — tool-call descriptions (every tool-call card title, including run_code in PTC-style presets), thinking, and replies — each set to follow the web GUI language, force a specific language, or turn off (default: only descriptions, following the GUI). One global runtime-context prompt directive on the host; the client half reports the GUI language; a Settings plugin card switches the modes.

OpenAI-completions-compatible adapter for custom gateways (vLLM / LM Studio / self-hosted proxies): always role:"system", thinking driven by the model config, Qwen-style response split, and vision-model image input (single / multiple).

Rule-based multi-provider model routing for DeepSeek Harness with pre-first-token failover, cooldown circuit-breaking, usage accounting, and a status API.

Per-provider concurrency gate on the llm/stream waterfall: holds surplus model requests in a FIFO queue inside DSH so main agent, subagents and compaction take turns on a single-slot local server (llama.cpp --parallel 1) instead of timing out; maxConcurrent / maxQueued / queueTimeoutMs.

One-click/auto sync provider model lists against /v1/models endpoint.

Explicit model routing for Main, Subagents, web_search, and image generation, plus a catalog-driven Composer and Plan Review picker with Model, Effort, Context, Fast, and Thinking variants.

Global Alt+M model command palette for DeepSeek Harness Web: fuzzy search over models and providers, favorites and recents, per-model reasoning effort, provider and model configuration, request retries, capacity presets, and optional OpenRouter image/video generation tools.

Provider/model-aware adaptive pacing and visible rate-limit retries for DeepSeek Harness free models.

在模型座左侧显示当前模型所属提供商（显示名称优先，回退 Provider ID）。纯运行时装饰，不改动原生页面源码。

让 DeepSeek Harness 自动重试商汤 429/QUOTA 等瞬时网关故障直到任务完成——启动时给所有 llm-pi-ai provider 热注入加固版 retryPolicy。

Free-tier model aggregation for the DeepSeek Harness (dsh): registers the freeroute provider over a pool of free-quota upstreams (OpenCode Zen / B.AI / OpenRouter / SenseNova built in, plus remote catalogs and custom gateways), transparent failover before

Run the main session model on a local Claude Code subscription through the official Claude Agent SDK, taking over the llm/stream route for the claude-code provider, with token-level streaming, session resume, and DSH tools bridged in over MCP so they keep DSH sandbox and approval.

Manages the default DSH model, provider and model visibility, and optional OpenAI Responses web search.

Planner/executor delegation with named roles: a strong planner steers, disposable executor sub-agents (per-role routes and prompts) implement. Default role plus per-delegation override, a global reasoning-effort knob in the composer chip, and a settings page for role management.

Adaptive model discovery, evaluation, tiering, and subagent routing for DeepSeek Harness.

Single-model disable gate: hides denied models from DSH catalogs, blocks direct dispatch with an in-protocol error chunk, and exposes a settings UI for the denylist.

DeepSeek Harness bridge plugin: exposes the newest Zhipu/Z.AI coding-plan models (e.g. glm-5.3) on the zai-coding-cn route before the bundled pi-ai catalog catches up.

Connect a stock DeepSeek Harness install to DSH Cloud: device login plus a 20-model gateway provider. Commercial hosted service by AgentsDance AI — new accounts include 500 free credits.

Reasoning-effort tier manager for custom API models plus cross-session preference memory: inline low/medium/high/max toggles inside Settings → Models, with atomic settings writes.

Free-tier model routing for the DeepSeek Harness (dsh) - discover OpenRouter free models, apply the best free model as agent default, auto-rotate on rate limits; paid config untouched unless asked.

Use ChatGPT (Codex) and Gemini Code Assist subscriptions as DeepSeek Harness LLM providers — OAuth login in the web UI, no API keys

DSH web search and Responses native V2 compaction for Sub2API-proxied or NewAPI-relayed GPT models

Hybrid model router for DeepSeek Harness: declarative rules, user-configurable cost control (cost-first / quality-first / balanced), optional LLM task classifier, and failure fallback chains.