Claude Skills by tody-agent

Use when starting any new project from scratch. Asks for project identity (name, GitHub org, Cloudflare account), detects project type, sets up design system, staging+production, i18n from day 1, SEO foundation, AGENTS.md manifest, test infrastructure, 8-gate deploy pipeline, and disciplined development workflows. Prevents wrong deploys, redundant repos, and technical debt from day 0.

Use at end of sprint to append operational learnings as structured JSONL to the project retro log.

Skill Chain Engine — compose skills into automated pipelines. One task triggers multi-skill workflows with progress tracking, auto-detection, and step management.

Progressive Disclosure skill index — efficient skill loading in 3 layers. Layer 1 (index, ~100 tokens) always loaded. Layer 2 (summary, ~300 tokens) loaded on context need. Layer 3 (full SKILL.md) loaded on execution only.

Easy-to-use conversational CLI (Claude Code style) for non-technical users to spawn parallel AI tasks supervised by a visual web dashboard.

Working memory protocol — maintains context across sessions via CONTINUITY.md. Inspired by Loki Mode. Read at turn start, update at turn end. Captures mistakes and learnings to prevent repeating errors.

Track and manage tasks across multiple AI coding agents (Codex, Google Antigravity, Cursor, Gemini CLI, Claude Code) through the Mission Control web UI at http://localhost:6969. Use when you need a visual kanban for multi-agent work, dispatching tasks, monitoring agent activity, or syncing task state from a CLI agent.

Optional power-up — detects oversized codebases/docs and suggests tobi/qmd for local semantic search. Bridges cm-continuity (working memory) with long-term document retrieval. Zero-config detection, non-intrusive suggestion.

A template demonstrating the universal format compatible with Antigravity, Claude, Cursor, and OpenClaw.

Use when executing approved implementation plans. Routes to 1 of 6 execution modes (batch/subagent/parallel/RARV/TRIZ/party) based on task shape. Loads detail refs on-demand.

You MUST use this before any creative work or multi-step task. Explores intent and writes execution-ready plans before coding.

Use when starting any new project from scratch. Asks for project identity (name, GitHub org, Cloudflare account), detects project type, sets up design system, staging+production, i18n from day 1, SEO foundation, AGENTS.md manifest, test infrastructure, 8-gate deploy pipeline, and disciplined development workflows. Prevents wrong deploys, redundant repos, and technical debt from day 0.

Use at end of sprint to append operational learnings as structured JSONL to the project retro log.

Skill Chain Engine — compose skills into automated pipelines. One task triggers multi-skill workflows with progress tracking, auto-detection, and step management.

Progressive Disclosure skill index — efficient skill loading in 3 layers. Layer 1 (index, ~100 tokens) always loaded. Layer 2 (summary, ~300 tokens) loaded on context need. Layer 3 (full SKILL.md) loaded on execution only.

Easy-to-use conversational CLI (Claude Code style) for non-technical users to spawn parallel AI tasks supervised by a visual web dashboard.

Working memory protocol — maintains context across sessions via CONTINUITY.md. Inspired by Loki Mode. Read at turn start, update at turn end. Captures mistakes and learnings to prevent repeating errors.

Optional power-up — detects oversized codebases/docs and suggests tobi/qmd for local semantic search. Bridges cm-continuity (working memory) with long-term document retrieval. Zero-config detection, non-intrusive suggestion.

A template demonstrating the universal format compatible with Antigravity, Claude, Cursor, and OpenClaw.

Use when executing approved implementation plans. Routes to 1 of 6 execution modes (batch/subagent/parallel/RARV/TRIZ/party) based on task shape. Loads detail refs on-demand.

You MUST use this before any creative work or multi-step task. Explores intent and writes execution-ready plans before coding.

Use when starting any new project from scratch. Asks for project identity (name, GitHub org, Cloudflare account), detects project type, sets up design system, staging+production, i18n from day 1, SEO foundation, AGENTS.md manifest, test infrastructure, 8-gate deploy pipeline, and disciplined development workflows. Prevents wrong deploys, redundant repos, and technical debt from day 0.

Use at end of sprint to append operational learnings as structured JSONL to the project retro log.

Skill Chain Engine — compose skills into automated pipelines. One task triggers multi-skill workflows with progress tracking, auto-detection, and step management.

Progressive Disclosure skill index — efficient skill loading in 3 layers. Layer 1 (index, ~100 tokens) always loaded. Layer 2 (summary, ~300 tokens) loaded on context need. Layer 3 (full SKILL.md) loaded on execution only.

Easy-to-use conversational CLI (Claude Code style) for non-technical users to spawn parallel AI tasks supervised by a visual web dashboard.

Strategic analysis gate for existing products — multi-dimensional evaluation (tech, product, design, business) using Design Thinking + 9 Windows (TRIZ) + Double Diamond. Outputs 2-3 qualified options with recommendations. Use BEFORE cm-planning for complex initiatives and enhancements on existing codebases.

Full review lifecycle — request reviews, handle feedback with technical rigor, and complete branch integration. Use when completing tasks, receiving feedback, or finishing feature branches.

Unified code intelligence — routes to Skeleton Index, CodeGraph, Architecture Diagram, or Smart Context Builder based on task shape. Loads deep refs on demand.

Working memory protocol — routes to session protocol, memory audit, MCP/URI integration, or storage details based on the current need. Keeps default continuity usage cheap.

Optional power-up — detects oversized codebases/docs and suggests tobi/qmd for local semantic search. Bridges cm-continuity (working memory) with long-term document retrieval. Zero-config detection, non-intrusive suggestion.

Ultimate Design System Intelligence. Manage, extract, replicate, and improve UI design systems. Extracts design tokens visually from existing sites (Harvester v5), manages multi-page tokens, and utilizes MCP servers like Google Stitch and Pencil.dev to generate complete, consistent design systems (Shadcn UI, Halo, Lunaris, Nitro). Use when the user asks to "extract design from a URL", "create a Shadcn design system", "analyze design tokens", "build a design system for", or "copy the UI style ...

A template demonstrating the universal format compatible with Antigravity, Claude, Cursor, and OpenClaw.

Use when executing approved implementation plans. Routes to 1 of 6 execution modes (batch/subagent/parallel/RARV/TRIZ/party) based on task shape. Loads detail refs on-demand.

Complete guide to vibe coding with the CodyMaster skill kit — from idea to deploy. Covers the full workflow, skills used at each phase, and common use cases. Read this first if you are new; reference it whenever you're unsure which skill to invoke.

Verify and lock project identity before ANY git push, Cloudflare deploy, or Supabase operation. Essential when working with multiple GitHub accounts (personal + work), multiple Cloudflare accounts, or multiple Supabase/Neon projects. Prevents wrong-account deploys, cross-project secret leaks, and git history contamination.

You MUST use this before any creative work or multi-step task. Explores intent and writes execution-ready plans before coding.

Use when starting any new project from scratch. Asks for project identity (name, GitHub org, Cloudflare account), detects project type, sets up design system, staging+production, i18n from day 1, SEO foundation, AGENTS.md manifest, test infrastructure, 8-gate deploy pipeline, and disciplined development workflows. Prevents wrong deploys, redundant repos, and technical debt from day 0.

Use before any deployment or completion claim. Enforces test gates, evidence-based verification, and frontend safety checks. No deploy without passing. No claims without evidence.

Use at end of sprint to append operational learnings as structured JSONL to the project retro log.

Use when setting up deployment infrastructure or enforcing a gated release path. Routes to deploy gates, setup, or rollback based on the current deployment stage.

Skill Chain Engine — compose skills into automated pipelines. One task triggers multi-skill workflows with progress tracking, auto-detection, and step management.

Progressive Disclosure skill index — efficient skill loading in 3 layers. Layer 1 (index, ~100 tokens) always loaded. Layer 2 (summary, ~300 tokens) loaded on context need. Layer 3 (full SKILL.md) loaded on execution only.

Start the CM Workflow to execute your objective from idea to production code.

Use when implementing any feature or bugfix. Routes to the core Red-Green-Refactor loop, test quality guidance, rationalization handling, examples, or stuck-state recovery.

Ultimate UI/UX design intelligence with Harvester v4 (AI-powered visual extraction), 48 UX Laws, 37 Design Tests, UX Heuristics (Nielsen + Krug), Figma & Google Stitch integration, MCP server for Claude/Cursor, Component Generator, and BM25 search across 16 domains. One command = Complete design system. 10x productivity boost.

Easy-to-use conversational CLI (Claude Code style) for non-technical users to spawn parallel AI tasks supervised by a visual web dashboard.

Working memory protocol — maintains context across sessions via CONTINUITY.md. Inspired by Loki Mode. Read at turn start, update at turn end. Captures mistakes and learnings to prevent repeating errors.

Optional power-up — detects oversized codebases/docs and suggests tobi/qmd for local semantic search. Bridges cm-continuity (working memory) with long-term document retrieval. Zero-config detection, non-intrusive suggestion.

A template demonstrating the universal format compatible with Antigravity, Claude, Cursor, and OpenClaw.