LLM Mart Basic

Micro-interactions are small, purposeful animations and responses that reward the user and make the interface feel alive — an animated icon, a satisfying toggle, a subtle reveal. Borrowed from the natural world, they add delight without distraction. Use when designing interactive

Overlays — modals, drawers, bottom sheets, popovers — interrupt or augment the main flow. Each type has a different scope, blocking level, and appropriate use case. Use when designing dialogs, confirmation prompts, side panels, action sheets, or any UI element that appears above

Typography feels cohesive and intentional when font sizes follow a modular scale — a ratio-based sequence where every size is mathematically related to the others. Use when defining type scales, setting up design tokens, reviewing font size choices, or when typography feels incon

Disney's 12 animation principles, cinematic storytelling techniques, and comic book conventions apply to web UI — used subtly, they make interfaces feel alive, intentional, and emotionally resonant. Use when designing transitions, micro-interactions, onboarding flows, scroll anim

UI design and review should apply Nielsen's 10 Usability Heuristics — the foundational principles for evaluating and improving usability. Use when auditing an interface, designing interaction flows, writing error messages, or reviewing any UI for usability issues.

When something goes wrong, the user must be able to recover or try again. Toasts, inline errors, banners, and notification patterns each have a specific role. Use when designing error states, success confirmations, async feedback, in-place editing, or any system that communicates

Operational expert tools — used by domain specialists for hours every day — require a different design approach than consumer or occasional-use software. Information density, workflow linearity, and at-a-glance status take priority over whitespace and discoverability. Use when de

Audit UI performance with Lighthouse and fix Core Web Vitals — LCP, CLS, INP. Fast UI is good UX. Use when optimising page load, fixing layout shift, reducing input delay, improving Lighthouse scores, or reviewing images, fonts, and render-blocking resources.

UI patterns borrowed from the physical world feel immediately intuitive — cards feel graspable, carousels feel scrollable, drawers feel pullable. Use real-world metaphors deliberately to reduce the learning curve and make interactions feel natural. Use when designing layout patte

Any component rendered many times — cards, list rows, table cells, nav items, tiles, KPI widgets, feed entries — is a fixed slot model, not a free-form box. The same slots appear in the same place in every instance and stay aligned across siblings even when text and values vary i

Mobile, tablet, and desktop are different interaction paradigms — not the same layout scaled up or down. Sections can be hidden, repositioned, or made sticky on mobile. Navigation and primary actions move. Use when designing responsive layouts, adapting desktop UI for mobile, or

Scroll areas inside a layout should be avoided wherever possible. When unavoidable, allow only one scroll axis at a time and always keep the user in control. Use when designing layouts, data tables, panels, or any component that might introduce an inner scroll container.

Semantic HTML5, SEO fundamentals, alt texts, progressive enhancement, SPA considerations, device capability detection, and user context awareness. Good HTML is the foundation of accessibility, SEO, and resilient UI. Use when building any web UI, reviewing markup quality, or optim

Which CSS unit a value should be written in. rem for anything a text-size preference must move — type, control heights, padding around text. px only for what must not move — hairlines, borders, shadow offsets. Relative units for layout widths. Use when defining spacing and sizing

Keep status and error colours minimal and consistent — too many semantic colours confuse users. Each colour must mean exactly one thing. Errors should be recoverable, large failures must be prevented, and the UI should always give the user a path forward. Use when designing statu

Sticky and fixed positioning keeps critical UI persistent as the user scrolls — headers at the top, toolbars at the bottom on mobile. Use deliberately: too many fixed layers create visual noise and reduce content area. Use when designing navigation headers, bottom toolbars, float

Tabs organise related content under a shared context — switching tabs swaps the view without leaving the page. Use when a screen has multiple distinct content areas that share a common header or action set, and when the user needs to switch between them frequently. Use when desig

UI should make it immediately clear where the user is, what context they are operating in, and what their actions will affect. Use lines, regions, colour areas, breadcrumbs, and scope labels to communicate hierarchy and context — especially in deep navigation structures or multi-

UI density — how much information and how many features appear at once — should match the primary platform and user type. Desktop supports dense, feature-rich interfaces; mobile requires focused, reduced layouts. Enterprise power users tolerate higher density than occasional user

Related features and tasks — such as purchase flows, onboarding, or multi-step configuration — should be designed as natural, guided paths that feel coherent and fit the product hierarchy. Use wizards for complex sequential tasks. Use when designing flows, onboarding, checkout, s

Fourteen posts of being wrong in production, compressed to checkboxes

Healthy nodes, a quiet network, 300 restarts in three days, and a latency budget measured in milliseconds

Discovery worked. Ping worked. Every TCP connection timed out, and later the tunnel only worked when someone had a terminal open.

Every VM came back. The cluster did not. Declarative systems converge on config, and the datapath isn't config.

A surprising share of AI-in-the-terminal failures aren't the AI. They're zsh, and a version of bash from 2006.

A Claude Code plugin turns standalone project configuration into a namespaced, installable extension that teams and communities can update as one unit.

None of the safety came from the model. It came from six boring habits.

Skills package instructions and references. Subagents run work in a separate context and return results. They solve different problems and can be composed deliberately.

Six hours in, one step left, everything green, and the incident that didn't happen

CLAUDE.md carries persistent project context. Skills load reusable procedures when relevant. Separating stable facts from task-specific workflows keeps both easier to maintain.

Twenty minutes recovering secrets that never existed, and the one sentence from a human that ended it

An API request routing a model's tool call through an approval gate to a remote MCP server

31 config keys, two audits, and why the first one was wrong in both directions

The official MCP Registry stores standardized server metadata rather than package code. Publishers verify a namespace, describe installation or remote access, and submit immutable versions.

Everyone looks at the Dockerfile. The file that actually leaked the key was the project file.

Remote MCP authorization uses established OAuth standards, but secure integration still requires issuer validation, least-privilege scopes, protected token handling, and server-side enforcement.

"Copy it over and switch the reference" is two steps, and the outage lives in the one nobody checks

stdio fits local processes and prototypes. Streamable HTTP fits hosted services and shared integrations. The right choice follows where the capability runs and who must reach it.

The most important rule wasn't about what I could change. It was about what I was allowed to display.

Tools perform operations, resources expose readable context, and prompts provide reusable templates. Choosing the correct primitive makes an MCP server easier to understand and govern.

{{SKILL_ENTRY:refactor}}

{{SKILL_ENTRY:release}}

Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR.

Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or {{CMD:feature}}.

Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops.

Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes.

Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only.

Wire codeArbiter's statusline into ~/.claude/settings.json, or remove it.

The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md.

{{SKILL_ENTRY:security-architecture}}

{{SKILL_ENTRY:tribunal}}

Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges.

Copy a file OUT of a running sandbox box to the host — host-initiated egress only (docker cp). The reverse, a host→container bind, is impossible by construction.

Tear down a sandbox box — remove its container and named volume. --keep-volume leaves the volume; with no id, prune reclaims any leaked ca.sandbox=1-labeled object. Cached images are retained.

Run a single command inside a running sandbox box and capture a JSON result — exitCode, separate stdout/stderr, and a truncated flag past the byte cap. The scriptable exec seam.

Open an interactive shell inside a running sandbox box at /work/repo. Read-only root, non-root user, no host-FS access — explore the untrusted code interactively, then exit.

Pull an untrusted repo into an ephemeral, host-FS-isolated Docker container — clone into a named volume, build a dep-cached image, run under structural isolation. Network defaults to offline. Requires Docker and nixpacks.

Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs.

Inspect ADR health read-only; optionally select one ADR with --adr N.

Record user-decided ADRs or inspect their health read-only. Preserve attribution and acceptance evidence.

Okou connects to the tools your team already uses and does the work — across marketing, sales, engineering, and operations, under your control.

An open-source Digital Worker platform for reliable execution, continuous co-evolution, and building Enterprise AI assets.

A curated list of AI Agent evolution, memory systems, multi-agent architectures, and self-improvement projects. | evomap.ai

A governance harness for AI coding.

DeepSeek Harness 手机版：可直接安装的 Android APK，AI 免 Root 操作手机（Shizuku/root 可选），文件编辑只需所有文件访问权限，前台保活 + AI 通知

Local-first, governed AI agent runtime for Python — embed it in your app, or run it as a CLI or ACP server. Permissions, MCP, memory and audit replay built in.

Orbi — the factory that builds and operates AI software factories. GitHub Issues in, releases and runnable system out

Open-source AI agent harness in native Rust — GUI, CLI, headless, and webapp from one binary. Multi-provider, MCP, skills, plugins, agent teams.

The SDK for browser agents. Interact, search, extract, and fetch any site reliably across the web

Agent-native shopping for extreme value: verifiable same-product price evidence, checkout, orders, delivery, and after-sales.

The extensible power-user platform for Google Antigravity. Adds a native In-App Browser, animated Desktop Pets, revamped Gemini UI, and custom BYOK Gemini Pro k…

Open-source, single-binary, self-hosted AI agent — your models and data stay on your machine. A coding agent on par with Claude Code and a personal assistant li…

Your pocket agent. Local-first AI agents on iOS and Android — real workspaces, tool execution with approvals, and your choice of model (DSH · Claude Code · Code…

Bridge between QQ (SnowLuma OneBot v11) and DeepSeek Harness agents: social simulation, safe MCP tools, slang learning and more.

A 7×6 framework for agent architecture. 28 patterns, each placed at a coordinate, runnable Python code with verified engineering slices from Claude Code, Aider,…

针对 AI 自动化渗透 Agent 的新一代反制蜜罐，通过反向代理将API密饵载入真实业务、反向提示词注入等方式反制自动化渗透 Agent，实现多款主流通用Agent的反制上线控制。

Open-source, local-first AI agent for coding and real work. BYOK models, MCP, skills, plugins, workflows, and private knowledge bases.

A free AI-agent toolbox for Android, 一站式安卓AI Agent工具箱

Google Workspace CLI — one command-line tool for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Dynamically built from Google Discovery Service. I…

Independent, unofficial CLI to edit CapCut and JianYing (剪映) projects — subtitles, timing, speed, volume, templates, cut long-form to shorts. No API needed, rea…