📚 Skill Archive

A comprehensive library of AI agent skills and cognitive frameworks, structurally organized around Lean manufacturing, Kaizen continuous improvement, and Toyota Production System (TPS) methodologies.

📂 Repository Structure

Directory	Purpose
skills/	High-level cognitive methodologies and tactical frameworks (Markdown + YAML frontmatter).
tools/	Low-level, standalone execution scripts (Bash/Python).
workflows/	Multi-step protocols for complex agent tasks.
templates/	Standardized formatting templates for new skills.

📋 Skill Index

🏗️ Architecture

Skill	Description	Tags
Interface Governance	Use when designing, building, auditing, testing, or consuming MCP servers and clients. Integrates API-Design-First pr...	architecture testing mcp methodology
Shusa	The Shusa (Chief Engineer) leadership model for technical ownership and product vision. Ensures that a single individ...	design leadership methodology lean TPS
Swarm Orchestration	Protocols for multi-agent coordination and delegation. Ensures that specialized sub-agents are tasked precisely and ...	communication design kubernetes
Value Stream Mapping	Use when diagnosing slow workflows, high token consumption, or unexplained latency in multi-step tasks. Handles curre...	design methodology optimization lean TPS

🧠 Cognition

Skill	Description	Tags
Comprehend	Use before debugging, extending, or shipping code. Acts as a strict policy gate implementing Risk-Tiered comprehensio...	cognition safety routing methodology
Deglaze	Use when reviewing AI-generated solutions, validating user comprehension, or stress-testing proposals. Handles 5 cons...	cognition safety methodology
Gemba	Use to establish a factual baseline of the codebase. Mandates direct observation before reasoning and enforces a stru...	context methodology research lean TPS
Hansei	Use when a plan fails, a bug is found, or a post-mortem is needed. Mandates 5-Whys root-cause analysis.	cognition methodology lean TPS kaizen

⚙️ Engineering

Skill	Description	Tags
Anchor	Use when starting a project, making architectural decisions, or when AI-suggested changes conflict with established p...	design architecture
Isolate	Use when a bug is reported, unexpected behavior occurs, or a fix attempt has failed. Handles 5-step hypothesis-driven...	debugging
PowerShell 7 Efficiency	Use when writing or optimizing PowerShell 7 scripts. Mandates the use of modern features like pipeline chaining, para...	engineering scripting automation performance cross-platform
Ship	Use when code is "feature complete" and approaching deployment, or when verifying production readiness. Handles stagi...	engineering safety methodology
Test-Driven Development	Use when implementing any feature or bugfix, before writing implementation code. Enforces the strict Red-Green-Refact...	engineering testing methodology
Verification Before Completion	Use when about to claim work is complete, fixed, or passing, before committing or creating PRs. Requires running veri...	safety engineering debugging

📌 Memory

Skill	Description	Tags
Context Engine	Use when required to retrieve information, learn codebase rules, read files over 4,000 tokens, or record short-term p...	memory mcp rag state

🔄 Meta

Skill	Description	Tags
External Synthesis	Use when researching new agent skills. Guides the agent to search external repositories and synthesize patterns into ...	meta research
Skill Authoring Management	Use when creating, reviewing, or managing agent skills. Provides the authoritative "Gold Standard" for content and f...	meta
Tools Management	Use when deciding whether to implement an abstraction as a cognitive Skill, an MCP Server, or a low-level Tool (scrip...	architecture design meta
Workflow Management	Use when creating, reviewing, or managing workflows. Ensures workflows meet the structural standards and prevents doc...	meta architecture

📐 Methodology

Skill	Description	Tags
Development Story Interview	Use when a user requests feature scoping, requirement definition, or planning a new idea. Applies Socratic questionin...	methodology lean cognition heijunka context-engine
Genchi Genbutsu	Use to "Go and See for Yourself." Mandates empirical verification of all hypotheses and findings through direct execu...	methodology research testing lean TPS
Heijunka	Use to level the workload and prevent token/context spikes (Muri). Handles the decomposition of massive tasks into m...	methodology optimization lean TPS
Hō-Ren-Sō	Use when reporting progress, broadcasting state changes, or escalating ambiguous or blocked decisions to a human oper...	communication methodology lean TPS
Incident Management	Use when creating, updating, resolving, or referencing incident reports. Handles the lifecycle of incidents (active/r...	debugging cognition context research
Kaizen	Use when a recurring error, structural bottleneck, or workflow inefficiency is identified (either manually or via Han...	methodology optimization lean TPS kaizen
Kodawari Craftsmanship	Use during any code modification or creation task as an always-active operating principle. Mandates the relentless pu...	design methodology lean TPS
Lean Foundations	Foundational Lean principles for stabilizing workflows through the 3 Ms (Mura, Muri, Muda) and the 5S framework. Focu...	methodology optimization lean TPS
Nemawashi	Use before suggesting major refactors or architectural changes. Mandates impact analysis and A3 proposals.	methodology lean TPS
Shisa Kanko	Use when executing code changes, tool calls, or multi-step tasks. Mandates pointing, calling, and verification.	methodology safety lean TPS cognition communication routing
Yokoten	Horizontal deployment of knowledge and best practices. Used to "broadcast" successful patterns or critical fixes fro...	methodology lean TPS kaizen

🛡️ Safety

Skill	Description	Tags
Jidoka (Autonomation)	Use when an abnormality occurs, an iteration threshold is breached, or before executing any action flagged by the Lay...	methodology safety lean TPS
KYT	Use before executing high-risk, destructive, or irreversible commands (rm, drop, reset). Mandates hazard prediction.	methodology safety lean TPS
Poka-yoke	Use when designing validation gates, enforcing schemas, or preventing invalid tool calls. Handles deterministic guard...	methodology safety lean TPS
Red-Teaming	Adversarial stress-testing for security and logic. Use to identify bypasses, edge-case failures, and "happy-path" bi...	safety security
Secure	Use when handling auth, user input, secrets, or API endpoints. Mandates trust-boundary audits.	security

📋 Workflow Index

Workflow	Description
Lean Analysis Workflow	Run a full Lean analysis on a document or topic using all 9 lean-tagged skills as analytical lenses
Kaizen Sprint Workflow	Standardized procedure for a rapid Kaizen iteration focusing on a single, isolated module.
TPS Architecture Review Workflow	Run a heavy-duty system-level architectural review for major refactors, cross-domain integrations, or organizational ...
AI Agent Skill Import	Procedure for synthesizing new agentic capabilities from an external repository into the local Skill Archive.
Master Topic Refinement Workflow (with Context Engine)	A multi-phase refinement protocol enhanced by the Context Engine to ensure a topic, documentation, or codebase reache...
Master Topic Refinement Workflow	A multi-phase refinement protocol to ensure a topic, documentation, or codebase reaches the highest level of Lean qua...
Parallel Processing Workflow (Map-Reduce)	High-speed execution pattern via parallel Map-Reduce, delegating chunks of work to concurrent sub-agents or determini...

🛠️ Tools Index

Standalone, low-level execution scripts. See the Tools Management Strategy.

Tool	Description
pycache
context-heijunka	This tool implements the Heijunka (Production Leveling) and Swarm Orchestration principles to prevent LLM context overburden (Muri) when analyzing external repositories.
docx-tools	A Python automation script using python-docx to read and generate Word documents programmatically.
git-worktree-setup	An automation bash script that streamlines the process of checking out a new feature branch in an isolated directory using Git Worktrees, followed by automated project setup.
image-converter
mcp-servers	The project is configured to use several Model Context Protocol (MCP) servers to extend the agent's capabilities with external tools, web interactions, and long-term memory.
pdf-tools	A Python tool leveraging PyMuPDF (fitz) for fast, robust PDF manipulation.
playwright-scaffold	A JavaScript automation script that injects a best-practice Playwright testing environment into any web project.
xlsx-tools	A Python script using openpyxl to extract data from Excel spreadsheets efficiently.
youtube-transcript	A Python automation tool using yt-dlp to fetch transcriptions and subtitles for YouTube videos.