Working Activity: Make the Status Line of DeepSeek Harness Truly Dynamic
working-activity is a community plugin maintained by chimney (ccch1mneyyy), which provides DeepSeek Harness with a real-time working status line driven by session events: thinking copy, running tools, time elapsed and closing summary. The same repository also maintains the pi CLI version, which is released as an independent npm package. This article verifies the installation commands, configuration items, TUI/dsh-cc/Web exits, and restrictions such as publish being disabled by default and no progress percentage against the directory page and GitHub. This is an unofficial project.
Read MoreUsing dsh-deep-whale to give DeepSeek Harness a whale girl skin
dsh-deep-whale is a DSH Web GUI whale girl skin series maintained by Small-tailqwq, and the current finished product is maid-atelier (Deep Sea Maid Workshop). This article verifies through the community directory page, GitHub README, skin.json and NOTICE that it only modifies the appearance of web pages and does not touch model requests; the installation command shall be subject to `dsh plugin add github:Small-tailqwq/dsh-deep-whale`; the license is CC BY-NC-SA 4.0, commercial use is prohibited, and the attribution chain must retain 上善, ZipZipPipe and the maintainer.
Read MoreConfigure Two-stage Agent Preset for DeepSeek Harness Using dsh-anchored-standard
*DSH-Anchored-Standard* is a preset for the community DeepSeek Harness maintained by xiaobright: In the first round, it aligns with the official Minimal trajectory using bash and str_replace_editor, and disables automatic injection. After a session generates persistable tool/call or assistant/message content, it will upgrade to a resident tool directory that can be unlocked on demand. This article cross-verifies the installation commands, copy enabling methods, compatible versions and experimental boundaries based on the community directory page and GitHub README.
Read MoreEquip Plain-text Coding Agents with a Visual Toolbox Using agent-vision-toolkit
Agent-Vision-Toolkit is an open-source visual toolkit (MIT License, Python) maintained by Anionex. It provides skills such as glance/ground/detect/trace/crop and vision-tools for pure text-coding intelligent agents like DeepSeek, covering visual question answering, long-screenshot OCR, UI restoration and GUI operation, and can be connected to Codex, Claude Code and other models. The community directory installation command points to this repository; the dsh Web/Headless native Bundle requires additional installation of @anionex/dsh-vision-toolkit.
Read MoreAdd a Sidebar Workbench to the DeepSeek Harness Web Interface Using DSH-better-sidebar
DSH-better-sidebar is a web interface enhancement plugin for DeepSeek Harness maintained by omdsh-dev. It provides file editing preview, real terminal, Git and sub-agent workbench in the right sidebar and bottom panel, and opens the ctx.betterSidebar registration interface for other plugins. This article verifies the installation command, the capability boundaries and precautions of v0.12.3 based on the community directory page and GitHub README.
Read MoreAdd Full-Screen Terminal Interface to DeepSeek Harness with dsh-TUI
dsh-TUI is a full-screen terminal plugin for DeepSeek Harness (MIT licensed) maintained by ccch1mneyyy. It is hooked into the dsh-tui profile via Cordis without modifying the core, and provides Claude Code-style whale top bar, status line, streaming thinking, double-press Esc to roll back, context progress bar and TPS meter. The directory command is `dsh plugin add github:ccch1mneyyy/dsh-TUI`, and it can also be installed via npm and run with `dsh-tui`. It is suitable for SSH/tmux. The plugin runs with the current dsh permissions, and you should check the source code and license before installation.
Read MoreAdd Vision Capabilities to the Plain-text Models of DeepSeek Harness with ModLens
ModLens is a DeepSeek Harness visual plugin maintained by Liustack (MIT License, version 3.18.1 at the time of writing). When plain-text DeepSeek/GLM cannot recognize screenshots, it registers the `modlens_read_image` function to convert pasted images into JSON containing OCR results, layout information and semantic data. This article verifies the installation command, two image pasting paths, engine configuration and output contract according to the community directory and GitHub.
Read MoreUsing dsh-web-ui to turn DeepSeek Harness web interface into a workbench
dsh-web-ui is a collection of DeepSeek Harness Web UI plugins and skins maintained by zhu1090093659, which is open-sourced under the Apache-2.0 license. This article checks against the community directory page, GitHub README and npm aggregate package to explain how the task board, Git graph, right panel, mobile remote, SSH, image understanding and skin center work; the difference between the directory installation command and @linxin666/dsh-web-ui-all; as well as restrictions such as permissions, cron scheduling and pnpm access control.
Read MoreInstall a RN Performance Optimization Manual for AI Assistant with react-native-best-practices
Callstack split the RN performance e-book into Agent Skill react-native-best-practices, which covers FPS, TTI, package volume, memory and native modules. This article is verified against the official SKILL.md and the repository README: 29 references cover how to triage, how to install the skills CLI, as well as list virtualization, React DevTools measurement and the usage of Android R8. The core is to measure before making changes.
Read More### 贴合技术文档场景的标准译法： `vercel-cli-with-tokens: Authenticate Vercel CLI via Tokens, No More `vercel login` Required`
`vercel-cli-with-tokens` is one of the Skills in Vercel's official agent-skills collection. It teaches Agents to call the Vercel CLI using the `VERCEL_TOKEN` instead of `vercel login`. It discovers tokens via environment variables and `.env` files, uses `VERCEL_ORG_ID` and `VERCEL_PROJECT_ID` in pairs, defaults to preview deployments, and supports overriding environment variables, build logs and domain names. It is suitable for CI and non-interactive automation scenarios, and complements `deploy-to-vercel`. It is forbidden to write Tokens into the `--token` parameter.
Read More# Vercel React View Transitions: Adding Native Transition Animations to React Following the Official Workflow
`vercel-react-view-transitions` is an official Agent Skill from Vercel Labs, teaching Cursor, Claude Code and Codex to use React's ViewTransition and `addTransitionType` to implement native page transitions. It can be installed via `npx skills add`. It covers shared element transitions, Suspense reveal, list reordering and route direction animations, with instructions for Next.js App Router included. ViewTransition is currently available in the React Canary channel; App Router has built-in support and no longer requires installing `react@canary`. Unsupported browsers will simply not display the animations.
Read MoreAudit the documentation against Vercel writing guidelines using writing-guidelines
This is an introduction to Vercel Labs' Agent Skill "writing-guidelines": it reviews the tone, structure, readability and typesetting of documents and prose in accordance with the Vercel Writing Guidelines. The following details are covered: its attribution and version, the mechanism of pulling the remote `command.md` before each review, the coverage of over 80 rules, the `file:line` output format, as well as the installation and triggering methods using `npx skills add` in tools such as Cursor, Claude Code and Codex, plus notes including network dependencies and Vercel site-specific conventions.
Read More# vercel-optimize: Optimize Cost and Performance of Vercel Projects Based on Production Metrics First
`vercel-optimize` is an official Agent Skill from vercel-labs/agent-skills (SKILL.md 1.2.0) that performs cost and performance audits for deployed projects. The workflow involves first collecting metrics from the past 14 days, then identifying routes using a deterministic script, and finally modifying code. It has the most complete support for Next.js/SvelteKit, is available for Nuxt, and has limited support for Astro. The installation command is: `npx skills add vercel-labs/agent-skills --skill vercel-optimize`. You need Node.js 20+, Vercel CLI v53+, and have completed login/link operations; route-level sorting additionally requires Observability Plus.
Read MoreZero Skill: Automatically Discover and Invoke Paid External Tools on a Per-Use Basis When the Agent Says "Cannot Do It"
Zero is a search engine and payment layer designed for AI Agents. Official Skills will intervene when the model is about to say "cannot do" or ask users to register for APIs on their own, using zero search / get / fetch / review to discover and call external capabilities (x402 / MPP) on a pay-per-use basis. Users can log in with a unified wallet, eliminating the need to apply for individual Keys for each service. This article verifies the installation methods (Claude Code, Codex, Gemini CLI, Cursor) and cost control points based on the official repository, SKILL.md, zero.xyz and the npm package @zeroxyz/cli (version 1.30.0).
Read MoreVariant Analysis: Locate a Bug and Eliminate All Similar Ones
Trail of Bits’ variant-analysis has turned the daily variant analysis work of security engineers into an Agent Skill: after confirming a vulnerability, it extracts the root cause, calibrates with precise patterns, searches the entire codebase while relaxing only one condition at a time, and finally triages and leaves CI rules. This article explains the five-step workflow, the selection of ripgrep, Semgrep and CodeQL, as well as the installation and invocation methods of Claude Code, Codex and Cursor, based on the official SKILL.md, plugin README and policy documents.
Read MoreSemgrep-Rule-Creator: Using Agent to Write Testable Semgrep Security Rules
`semgrep-rule-creator` is an Agent Skill launched by Trail of Bits, used to create custom Semgrep rules that can detect security vulnerabilities, bug patterns and code patterns. Instead of running existing rule sets, it forces the Agent to produce rules following the workflow of test-first, AST analysis, priority given to taint mode, and optimization only after all tests pass. This article organizes the installation method, seven-step workflow, eval taint example and usage boundaries based on the official SKILL.md, plugin README and skill market description.
Read More# Property-Based Testing: Trail of Bits Turns Property Testing into Reusable Agent Skills
Trail of Bits' property-based-testing is a cross-language skill guide for property testing that includes smart contracts. It teaches agents to identify encoding/decoding, parsing, normalization, and contract invariants, write tests based on properties like round-trip and idempotence, and distinguish between test errors and real bugs when tests fail. Based on the official SKILL.md, plugin READMEs and market repositories, the article explains the installation of Claude Code / Codex, the `npx skills add` command, and reproducible examples such as Hypothesis.
Read MoreDifferential Review: Trail of Bits formalizes PR security differential review into an executable workflow
Differential-review is an Agent Skill released by Trail of Bits, which conducts security-focused reviews on PRs, commits and diffs: it classifies tasks by risk rather than line count, uses git history to detect regressions, quantitatively calculates the blast radius, treats missing tests as a higher risk factor, and enforces the output of a Markdown report with line numbers and attack scenarios. This article checks the official SKILL.md, plugin README and documentation website, explaining what problems it solves, how to install it in Claude Code / Codex / the general Skill catalog, as well as three official usage scenarios and inapplicable scenarios.
Read MoreUsing firecrawl-build-search: Start from queries and write web discovery into product code
`firecrawl-build-search` is an official Agent Skill for Firecrawl, specifically designed to connect the `/search` endpoint to product code: the functionality starts from queries instead of URLs. It first discovers, ranks and selects sources, then decides whether to hydrate the content or hand it over to `/scrape`. This article cross-verifies installation commands, authentication fields, one-step/two-step call differences, and error-prone upgrade paths such as paper indexing and developer indexing based on GitHub's SKILL.md, repository README, officialskills.sh and docs.firecrawl.dev.
Read MoreIntegrate Web Search, Crawling and Browser Interaction into Application Code with firecrawl-build
`firecrawl-build` is the official application integration entry Skill of Firecrawl, located in the `firecrawl/skills` repository. It targets scenarios where products, agents or workflows need to obtain webpage data through code. It first clarifies what data to fetch and how to fetch it, then routes the requirements to `/scrape`, `/search` or `/interact`, instead of mixing "help me search a page in the current session" with "integrate Firecrawl into the backend". This article checks the installation commands, environment variables, SDK, sub-skill division, smoke test, and the boundaries with CLI/Workflow skills against the SKILL.md, repository README and official documentation.
Read MoreSandbox SDK Skill: Let AI Securely Execute Untrusted Code with Cloudflare Sandbox
### Official Translation: This article introduces Cloudflare's sandbox-sdk Agent Skill: it explains how to guide Cursor / Claude Code / Codex to use the Sandbox SDK to execute untrusted code in Workers isolated containers. The article verifies the current status of the project being split into sandbox-stable, sandbox-next, and sandbox-migrate-to-next in August 2026, and provides official installation methods, minimal Worker examples and lifecycle, as well as restrictions such as transport and image alignment.
Read More# workers-best-practices Skill: Review and Write Cloudflare Workers in Accordance with Production Agreements
Cloudflare's official workers-best-practices Skill is used to review and write Workers code in accordance with production conventions: it pulls current documentation and types first, then checks `wrangler.jsonc`, streaming responses, floating promises, global state, secrets and observability. This article verifies the installation method, rule list and anti-patterns based on the original GitHub content and the official best practice page, and explains that Durable Objects, Workflows and the wrangler CLI need to switch to other Skills in the same repository.
Read More# 翻译结果 Durable Objects Skill: Build Chatrooms, Collaboration and Stateful RPC According to Official Conventions
### 正式翻译（符合技术文档译法，兼顾准确性与行业惯例）： The official Cloudflare durable-objects Skill is used to create and review Durable Objects: sharding by chat rooms, game sessions or tenants, routing via `getByName`, carrying stateful edge logic with SQLite, RPC, alarms and WebSockets, and testing with Vitest. This article verifies the installation, configuration and anti-patterns against the original GitHub source and official documentation, and notes that while the Skill example still uses migrations, the current getting-started documentation recommends declaring SQLite classes via `exports`.
Read More# Cloudflare Skill: A Progressive Loading Full Platform Manual for AI
Cloudflare's official cloudflare Skill is a cross-platform entry point for Agents: users can select the right primitive among Workers, Pages, KV/D1/R2, Workers AI, Network & Security, and Terraform/Pulumi via a decision tree, then load 63 product reference directories on demand. This article cross-verifies the installation method, binding syntax and typical usage based on the original GitHub text and official documentation, and notes that reference materials may expire — for example, the example model @cf/meta/llama-3.1-8b-instruct has been discontinued as of 2026-05-30, and the quota limits and model names shall be subject to the latest Cloudflare documentation.
Read MoreConnect 1000+ External Apps to Agents with Composio Skill
Composio's official Skill teaches AI programming assistants to connect to 1000+ external applications such as Gmail, Slack, GitHub, Notion via a unified CLI and SDK. This article verifies the installation commands, the search→link→execute workflow, user-isolated sessions, as well as the usage boundaries of OAuth and Triggers based on SKILL.md, the repository README and docs.composio.dev.
Read More