skills.rest
NAM-MAN
NAM-MANCOMMUNITY
Explore all skills by this author →
GitHub

takumi

Manage end-to-end software workflows via natural language commands.

Updated Apr 19, 2026
Search Tags:#automation#testing#verification#workflow orchestration#telemetry#AI planning#project workflow
One-click install
npx skills add https://github.com/NAM-MAN/takumi --skill takumi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: takumi
Source: https://github.com/NAM-MAN/takumi/tree/main/skills/takumi
Command: npx skills add https://github.com/NAM-MAN/takumi --skill takumi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires internalRoles, telemetry, profiles, exec, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a single entry point to manage, analyze, and optimize complex AI-based development and review workflows, reducing fragmentation and manual overhead.

Core Features & Use Cases

  • Unified Command Interface: Uses /takumi <natural language> to automatically determine operational mode—plan, review, verify, or refactor—streamlining user interaction.
  • Task Planning & Execution: Generates high-quality, structured plans stored in .takumi/plans/ and auto-executes via internal roles, reducing manual orchestration.
  • Quality Assurance & Refinement: Integrates mutation testing, property-based testing, and model-based verification to ensure high software reliability with minimal manual intervention.
  • Workflow Optimization: Facilitates continuous review, drift detection, and progressive improvement with rich telemetry, managing complex multi-stage processes efficiently.
  • Cross-Model Verification & AI Review: Employs cross-model AI diagnostics and reviews to prevent blind spots, ensuring holistic problem detection.
  • Iterative Knowledge Evolution: Supports multi-phase, incremental refinement, including status checks, pauses, and resumes, to adapt to project dynamics seamlessly.
  • Internal Resource Management: Automates management of dependencies, environments, and components, integrating with codebases and project profiles for consistent operation.

Quick Start

Invoke the Skill with a natural language request such as "Generate a plan for the new payment module" to immediately initiate plan creation, review, and execution without manual toggling or multi-command workflows.

Frequently Asked Questions about takumi

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate end-to-end project workflow planning and testing in one place?▼

You can automate end-to-end project workflows by using a unified natural language command interface that handles planning, testing, verification, and review through internal orchestration roles and telemetry.

What is the best way to detect workflow drift during AI-driven software development?▼

Workflow drift is detected using continuous review and telemetry features that monitor multi-stage processes, enabling progressive improvement and iterative knowledge evolution throughout the software lifecycle.

Can I use natural language commands to manage project refactoring and verification?▼

Natural language commands can manage project refactoring and verification by automatically determining the operational mode, streamlining user interaction without manual toggling or multi-command workflows.

How does cross-model AI review work for preventing blind spots in code verification?▼

Cross-model AI review employs multiple AI diagnostics to prevent blind spots and ensure holistic problem detection, cross-checking outputs across different models during the verification and review phases.

Do I need internal resource configurations for workflow orchestration and telemetry?▼

Internal resource configurations are required to manage dependencies, environments, and profiles, ensuring consistent operation when integrating workflow orchestration and telemetry with your existing codebase.

When should I not use a unified task hub for software lifecycle automation?▼

You should avoid a unified task hub for software lifecycle automation if your project lacks the required telemetry and internal role dependencies, as these are strictly necessary for continuous process improvement.

Related Skills

Explore More Software Engineering Skills→
KrishpotanwarKrishpotanwar

swarm-advanced

Orchestrate multi-agent swarms for distributed workflow management.

Community
Advanced
CAPHTECHCAPHTECH
1

eld-ground-law-monitor

Detect production-law violations from telemetry and log events, then upsert findings to pce-memory.

Official
Advanced
ThejanaJayalathThejanaJayalath

kotlin-testing

Standardize Kotlin testing workflows with Kotest, MockK, and Kover.

Community
Advanced
thomfilgthomfilg

check

Automate code review, test execution, and requirements verification.

Community
Intermediate