模型与账号

帮助你的 agent 尝试另一种方法，并在宣布完成前验证其工作。

Help your agent try another approach and verify its work before calling it done.

TinyFish-backed search and fetch providers for the DeepSeek Harness web capability seam (ctx.web) — $0 SERP and page extraction, direct or via Monid.

自动复盘插件：对话收尾后自动复盘，把有价值的经验蒸馏成可复用的技能（skill），把值得长期记住的事实与偏好写入跨会话记忆。

dsh 插件 · 自动复盘：对话收尾后自动复盘，把有价值的经验蒸馏成可复用的技能（skill）存入技能库，模型可直接调用；同时把环境事实与用户偏好写入跨会话记忆（MEMORY.md / USER.md，首冻结注入每个对话）；支持信号加速触发、手动立即复盘、审批模式与技能库治理（归档/恢复，永不直接删除）。

DeepSeek Harness × Google Antigravity — 多账号智能池化与官方 agy CLI 驱动桥接 / Multi-Account Pool & Native agy CLI Bridge for DSH。

Google Antigravity (agy CLI) models for DeepSeek Harness — stream Gemini/Claude/GPT-OSS subscriptions into DSH with thinking, tool activity, token usage and in-GUI Google OAuth login.

为 DeepSeek Harness 生成的模型会话标题能够实际运行、跟随你的最新任务，并报告自身的失败。

Model-generated session titles for DeepSeek Harness that actually run, follow your latest task, and report their own failures.

dupguard 是 DeepSeek Harness (DSH) 的实时重复输出守卫插件：最新输出中同一字符串连续重复 10 次及以上（可配置）时立即停止本次生成，已生成内容正常提交为助手消息。

Real-time repetition guard for DeepSeek Harness (DSH): stops generation once the same string repeats >=10 times in a row, with one relaxed threshold for fenced/inline/indented code, table- or module-driven repeat counts, and a derived detection window wri

我的任务，要去哪里？？？？？

My tasks, where to GO?????

DSH 的 OpenCode GO provider，其模型列表与能力为实时读取，而非取自 DSH 包内附带的快照——正因如此，单一路由即可服务网关提供的所有模型，并覆盖其全部 wire 协议。

An OpenCode GO provider for DSH whose model list and capabilities are read live instead of from a snapshot shipped inside the DSH package — which is what lets one route serve every model the gateway offers, across all of its wire protocols.

dsh-opencode-patch 是一个 DeepSeek Harness 宿主插件，用于让 OpenCode Zen 与 Go 模型在 DSH 中保持可用。

dsh-opencode-patch is a DeepSeek Harness host plugin that keeps OpenCode Zen & Go models working inside DSH.

一个 DeepSeek Harness (DSH) 客户端插件，在对话视图中添加“执行图”标签页。

A DeepSeek Harness (DSH) client plugin that adds an "Execution graph" tab to the conversation view.

DSH 模型治理平面：按提供商进行 RPM 排队（延迟而非拒绝）、RPS 爬坡式 RPM 探测（仅在达到上限的运行中自动应用），以及 OpenCode 会话头兼容。

DSH model-governance plane: per-provider RPM queueing (delays, never rejects), RPS-ramp RPM probing that auto-applies only on topped runs, plus OpenCode session-header compat.

在 DeepSeek Harness 中，把你已付费的 AI 订阅用作 LLM 提供方——无需 API key。

Use the AI subscriptions you already pay for as LLM providers in DeepSeek Harness — no API key.

Weak-network adaptor for the DeepSeek Harness — 弱网适配插件。

Weak-network adaptor for the DeepSeek Harness: transparent long-backoff model-stream retries (10-min cap), local response replay cache, heartbeat auto-reconnect, degraded-mode token economy, and a bilingual (zh/en) settings page. / 弱网适配插件：模型流长退避重试、本地响应缓存、

DeepSeek Harness Web GUI 的提示词增强功能——一键将粗略的编写草稿转为结构良好的提示词：明确的角色与目标、可执行步骤、输出格式、验收标准以及边界情况。

Prompt enhancement for the DeepSeek Harness web GUI — one click turns a rough composer draft into a well-structured prompt: explicit role and goal, executable steps, output format, acceptance criteria, and edge cases.

DeepSeek Harness（DSH / cordis plugin）插件：通过 Jev 或工具调用，让 Agent 根据任务场景自主切换模型预设和思考深度。

A DeepSeek Harness (DSH) plugin: through the TypeSafe Jev decision model (it reads each turn's instruction, classifies the task and picks a preset) or through its own tool calls, the agent switches model presets and thinking depth on its own, following th

Telegram channel for DeepSeek Harness — rich tables and lists, real image delivery both ways, inline-keyboard questions, and in-chat model/preset/mode control

可移植的设置传输——插件现在会在 settingsScope（≤ 0.1.5）和 configForms（≥ 0.1.7-rc.1）两种宿主上解析其设置作用域，因此 Settings 卡片和 pill 开关在重命名后仍能正常工作。

Portable settings transport — the plugin now resolves its settings scope on both the settingsScope (≤ 0.1.5) and configForms (≥ 0.1.7-rc.1) hosts, so the Settings card and pill switch keep working across the rename.

在 DeepSeek Harness（dsh）的 Web 界面里，在每个已完成回合的下方单独一行显示这一轮花了多少钱，点击可展开明细（各 token 类别的数量 × 单价 = 金额、命中的价格规则、计费时刻）。

DeepSeek Harness (dsh) plugin: shows the CNY cost of every completed turn under the turn, from per-model cache-hit / cache-miss / cache-write / output rates, with weekly peak-valley multipliers and limited-time promo windows.

DeepSeek Harness 的持久化提供商上报 token 使用统计。

Persistent provider-reported token usage statistics for DeepSeek Harness.

在 Web UI 中为当前打开的对话显示实时 USD 费用徽章的树外 DSH 插件。

Out-of-tree dsh plugin that shows a live USD cost badge for the open conversation in the Web UI.

DeepSeek Harness (DSH) 原生无感记忆与自省管理插件。

Cross-session and cross-project memory manager for dsh (silent turn-end sedimentation + native web client embedding)

DeepSeek Harness (DSH) 插件：在设置页提供「模型健康」面板——列出所有已配置模型，并一键批量测试可用性与延迟。

DSH plugin that lists all configured models with availability/latency testing, by reading $DSH_HOME/settings.yaml.

俄语指南：Jira 和 Confluence 的配置。

Русский гайд: настройка Jira и Confluence。

DSH 费用统计插件:会话费用行 + 成本归因报告,llm-pricing 动态价格。 / DSH cost attribution and usage dashboard, priced by llm-pricing.