improve-claude-md
claudeRewrite a CLAUDE.md file using <important if> blocks to improve instruction adherence
1 skill
Claude Code and Codex plugins their authors filed under "productivity". All categories.
Rewrite a CLAUDE.md file using <important if> blocks to improve instruction adherence
1 skill
Optional add-on for Claude Code: Session Orchestrator status line, warnings and session beacons, plus protection gates (secret-printing commands refused; pushes to main, merges, releases, publishes and deploys confirmed first). Needs the Session Orchestrator release that ships it, or later.
Agentic SEO Skills with project brains, DataForSEO workflows, content, and technical SEO.
21 skills · 2 commands · 13 agents · hooks
Fourteen Claude skills that run an Instagram account: a scored hook panel, a viral research skill, a caption linter, a humanizer, and publishing, comment replies, comment-to-DM automations and analytics through the PostZen MCP server. Nothing is posted until you say yes.
14 skills · 1 MCP server
Memory redirect hooks, session capture, project bootstrap, and maintenance skills for rekal.
4 skills · hooks
Local RAG (Retrieval-Augmented Generation) system for enhanced Claude Code responses with multi-agent orchestration
1 skill · 1 MCP server
覆盖软件开发每个阶段的生产级工程 skills:spec、plan、build、verify、review 和 ship。
26 skills · 4 agents
Dashboard-centric Agent Project OS with project bootstrap, execution loops, audit visibility, validation, reporting, and executor-mode Codex delegation.
1 skill · 11 commands
Eleven Claude skills that get you paying customers before you launch: a market researcher, a demo builder, a whale hunter, a pre-sale skill, requirements, an offer builder, pricing, follow-ups, an objection handler, case studies and a launch planner.
11 skills
Design and build a finished motion piece (vertical 1080×1920 by default; landscape/square supported) from a brief. A fresh-context concept step lists N directions that differ at the idea level and each of N independent draws is dealt one; a unified design+build agent (乙) designs the piece and writes the Remotion code in one continuous context and renders a first full cut (the r1 preview); the most promising base is picked from those previews — by the user by default, or by a blind selector; only the picked draw self-checks. Polishing is the 甲乙环 critic loop by default — a design-blind aesthetic critic (甲) judges the rendered video, its time overview and settle frames round after round until it converges: hands-off, safe, not ceiling-breaking — or, when the user chooses 亲自打磨, the user's own comments ferried verbatim to 乙. The user's own eyes are the final gate.
2 skills · 4 agents
Keep Codex and Claude inside clear boundaries when they read or update local Markdown notes.
5 skills
Hacker News for Claude. Scan the front page for today's themes, hot discussions, and industry mix.
1 skill
Инженерный skill для проектирования, реализации, отладки и оценки AI-агентов: архитектура, Harness, контекст, инструменты, память и evals.
1 skill
Codex plugin scaffold for SpringSentinel SpotBugs bytecode analysis.
1 skill
Generate Typst/Touying presentations from PDF / DOCX / URL / Markdown — 150+ themes, AI image generation, audio narration, and export to PDF/PNG/PPTX.
Generate HTML/SVG presentations with Satori rendering — 9 editorial templates, MathJax math, Prism.js code highlighting, and export to SVG/PNG/PPTX.
Memi is the design layer for agentic AI.
5 skills · 1 MCP server
Two skills: think-twice forces Claude to check for existing APIs, packages, and simpler approaches before implementing. Surgical forces Claude to build exactly what was asked, nothing more. Modeled savings: 178x on fake user profiles, 13x on currency conversion, 6x on city autocomplete, 5x on rate limiting.
3 skills · 2 commands
Checks every cut in a finished video: clicks, sync, frames, loudness and platform rules.
1 skill · 1 command · 1 MCP server
智能保存和恢复 Claude Code 会话,支持进度跟踪、决策记录和工作连续性,采用项目本地存储和智能数据访问设计
2 commands
Real-time status HUD for GitHub Copilot CLI. Shows project path, git branch, live context bar, tool activity, background agent tracking, and remote-session indicator.
2 commands
Filter noisy command output before it reaches the model; raw output stays recoverable. Skill + MCP server.
1 skill · 1 MCP server
Drive a real Chrome with four MCP tools costing 297 catalog tokens (Playwright MCP: ~13.7k). Compressed a11y snapshots, uid actions, CCR byte-exact recovery. Savings always inferred.
1 skill · 1 MCP server
Evidence-based resume tailoring and candidate profile onboarding. Review job fit, factual reliability and document quality separately; optionally export PDF or DOCX.
2 skills
基于 CLAUDE_TEMPLATE.md 模板快速初始化项目 CLAUDE.md 规范文件,通过交互式问答生成定制化开发规范
1 command
Maton
1 MCP server
Review ChatGPT Work conversations for durable lessons and export approved guidance without local hooks or background state.
1 skill
Complete project development toolkit: agents, reusable skills, and lifecycle skills (/spec to /ship) for Claude Code workflows
103 skills · 29 agents · hooks
Provider-agnostic AI skills for creating professional arc42 software architecture documentation. Covers all 12 arc42 sections plus cross-section quality review.
14 skills
Weekly summaries, standups, recall, and metrics for AI coding sessions across Claude Code, Codex, Pi, and OpenCode.
8 skills · 1 MCP server