Agent collaboration · DeepSeek Harness plugins — Page 253

SQL 优化模式：EXPLAIN 分析、索引策略、N+1 解决、查询改写。受 wshobson/agents（38k★ MIT）启发。

PickDOM：DSH 侧边栏双引擎 Browser，像 Cursor 一样框选本地 HTML 与 http/https 页面元素交给 Agent；本地文件支持可视化编辑并写回源码。

Route DeepSeek Harness model-API requests through an HTTP/SOCKS proxy without modifying Harness source

HTTP/API debugging toolset for DeepSeek Harness: http_request, curl_parse, request_history

古法编程模式（Classical Programming Mode）：AI 只出骨架、用户补全实现、AI 验收讲解的趣味编程练习模式。DeepSeek Harness 双面插件包（Host + Client）+ 随包 Agent 预设组合。

Human intent verification and cryptographic action authorization for AI agents. AI can propose an action. Only a human can authorize it.

Gives the DeepSeek Harness agent a configurable humanized whale-girl maid persona (address, self-name, speaking mode), an immersive no-plugin-meta rule, and a lightweight layered memory with query-triggered recall.

SQL 优化模式：EXPLAIN 分析、索引策略、N+1 解决、查询改写。受 wshobson/agents（38k★ MIT）启发。

IDE ↔ Harness 双向桥: IDE 侧 POST /ide-bridge {text} → 浏览器输入框草稿追加; 聊天内 /open-in-rider 链接 → Rider 跳转到对应行(toast 提示); /ide-bridge/links-config 两级开关(AGENTS.md 指令块)。

想法打磨：结构化发散-收敛思考，把模糊想法打磨成可执行的清晰概念。受 addyosmani/agent-skills（88k★ MIT）启发。

数字身份运维工具：邮箱（列/读/取码/取链）+ 站点知识注册表（可成长——新站点是加记录，不是加脚本）。纯 HTTP 经 Clash 显式代理，fail-closed。

Install iHow Memory as a local-first shared memory plugin for DeepSeek Harness.

Keep native vision, drop historical images before they hit relay providers. Global DeepSeek Harness bundle for every agent.

Give text-only models eyes: an analyze_image tool for DeepSeek Harness, backed by free Chinese vision APIs (GLM-4V-Flash / Qwen-VL) or any OpenAI-compatible vision endpoint. 给纯文本模型装上眼睛。

DSH 图片理解桥接插件:挂 agent/pre-step,把消息中的图片块经视觉模型(默认智谱 GLM-4V-Flash 免费档)转成文字描述,让纯文本模型(DeepSeek)也能看图

Inline previews for the images an agent already produced in DeepSeek Harness: read_image attachments and local paths in any format (HEIC/TIFF/PSD/SVG/BMP/ICO/TGA/PNM/JP2 …), rendered through the official gallery and lightbox.

Give DeepSeek Harness agents the ability to read images directly: a model-facing read_image tool that answers questions about an image through any OpenAI-compatible vision endpoint.

DeepSeek Harness plugin: an inline conversation image card with a bottom-left edit button (annotate / erase / resize) that hands the edit back to the agent.

DeepSeek Harness 原生插件：让纯文本模型（如 DeepSeek）经 agent/pre-step 劫持 + resolveModelInfo 包装自动识别上传图片（qwen-vl），使纯文本对话也能“看见”图片。

dsh 插件:图像生成/编辑/识图(imagegen + image_understand),provider 无关,默认提供方 Agnes AI(OpenAI 兼容)

Bundled imagegen skill for DeepSeek Harness: generate or edit raster images (photos, illustrations, textures, sprites, mockups, transparent-background cutouts) from the agent, through any OpenAI-compatible GPT Image endpoint.

dsh iMessage 插件：iMessage 收发完整接入 dsh — RPC 监听入站、按 handle 路由工作区、agent 处理自动回复、配置页（Typert remote）+ 全局 message 出站工具。随 dsh web 启停。

Load a workspace's own VSCode/Copilot AI configuration (.github/copilot-instructions.md, .github/instructions, .github/skills) into every DeepSeek Harness session. DSH 加载 VSCode AI 配置插件。

DSH plugin: sync chat.deepseek.com sessions into DSH as context — settings page, auto sync, per-session fetch, and a 'send' button that stages a file-reference prompt into the composer.