职场效率大师:用 Agent Skill 搭完整 AI 工作流

职场效率大师(@user_ef64daef/zhichangxiaolv,v1.0.1)是 SkillHub 社区 ai-agent 类 Skill,定位为 AI 工作流设计系统:覆盖诊断、方案、工具推荐、效果追踪与持续优化,并含 5 大场景、4 个职业深耕模块与 100+ 工具百科,标签 requires_api_key=false。本文基于 SkillHub 公开页与安装文档,说明其能力边界,给出 skillhub CLI 安装命令及 Cursor、Claude Code、Codex 等 skills 目录要点,并归纳可复现的对话用法原则与社区来源注意事项。

Read More
dsh-flomo-report: DeepSeek Harness Daily Report Generation and flomo Sync Tool

`dsh-flomo-report` is a workflow plugin for DeepSeek Harness (DSH) designed to automatically organize session records, generate AI summaries (daily and weekly reports, etc.), and push them to flomo. It supports session data extraction, Webhook delivery, custom configuration, and daily reminders. Users can install and use it via command, and should note the Node.js version and configuration permission requirements.

Read More
oh-my-dsh: Multi-agent Orchestration Plugin for DeepSeek Harness

oh-my-dsh is a community plugin for DeepSeek Harness (DSH) maintained by YYTbit, licensed under the MIT License, designed to provide multi-agent orchestration capabilities. This plugin decomposes development tasks into stages such as exploration, planning, implementation, review, and debugging, completing complex tasks through the collaboration of five specialized agents (Explorer, Architect, Implementer, Reviewer, and Debugger) and four workflow skills (Autopilot, Deepwork, Review, and Reflect). Core features include parallel execution, dependency tracking, runtime model configuration switching, and task workflow automation. Users can add oh-my-d via `dsh plugin --profile your-profile add oh-my-d`

Read More
dsh-mmx-bridge: Enabling DSH Access to MiniMax Multimodal Capabilities

The dsh-mmx-bridge is a DeepSeek Harness (DSH) MIT-licensed plugin maintained by welsione, designed to integrate MiniMax's multimodal capabilities into DSH agent workflows via the mmx_bridge tool. This plugin supports image understanding and generation, video creation, voice synthesis, music generation and cover production, online search, and quota queries. Users can issue direct commands in dialogue or drag and drop images for automatic recognition, with the recognition results featuring an embedded caching mechanism (stored in plaintext) for consistent queries on the same image. The plugin requires DSH version 0.1.0-rc.7 or higher and a logged-in mmx-cli environment. It can be installed via the `dsh plugin` command and takes effect after a restart. Generated artifacts are saved by default to `/tmp/mmx-out/` and are processed via the `/mmx` endpoint.

Read More
dsh-reference-anything: Integrating Multiple Reference Sources of DSH into an @ Menu

dsh-reference-anything is an MIT-licensed DSH plugin maintained by Chael-Chael, designed to address the issue of frequent switching caused by scattered multi-source data. This plugin integrates DSH commands, Skills, workspace files, session history, local Agent records, cloud drive files, and online chat history into the DeepSeek Harness @ menu, enabling one-stop search and reference. Its core functionalities include extending referenceable sources (supporting access to online sessions via OpenCLI and cloud drives via OpenList, with local records read directly from disk files) and enhancing menu interaction (supporting custom grouping order, folding settings, and optimized visual icons). The plugin is compatible with DSH versions 0.1.0-rc.8 and above, and after installation, it allows users to...

Read More
HN Hot Topic: Can Manually Re-typing Every Line of Code Generated by an LLM Avoid "Cognitive Debt"?

On August 4, 2026, Ankur Sethi published a post advocating for manually retyping LLM-generated code line by line to avoid "cognitive debt", which garnered 409 points and 348 comments on Hacker News. This article outlines its core workflow (the Agent only displays changes without directly writing files), the pros and cons debates on HN, as well as more widely accepted alternatives like plan-first, design-first, and test-driven development, to help developers balance efficiency and code comprehension in the era of AI programming.

Read More