skills.rest
tinner-deinno
tinner-deinnoCOMMUNITY
Explore all skills by this author →
GitHub

self-improve

Run PDCA cycles to store goal attempts as learning history.

1|Updated May 24, 2026
Search Tags:#workflow automation#self-improvement#multi-agent#pdca#knowledge persistence#goal execution#learning history
One-click install
npx skills add https://github.com/tinner-deinno/innova-skills-lib --skill self-improve-tinner-deinno
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: self-improve
Source: https://github.com/tinner-deinno/innova-skills-lib/tree/main/core/self-improve
Command: npx skills add https://github.com/tinner-deinno/innova-skills-lib --skill self-improve-tinner-deinno

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of repeating trial-and-error when pursuing a goal by turning each attempt into stored lessons, so future runs become more effective instead of starting from scratch.

Core Features & Use Cases

  • PDCA-based self-improvement loop: Plan, execute steps, check results, then Act by recording what worked and what didn’t.
  • History-driven strategy selection: Reuses past successful approaches and adapts when new goals resemble earlier ones.
  • Structured goal execution for complex tasks: Breaks a goal into smaller steps and runs appropriate sub-tasks (e.g., using helpers like learn/thread/taskcreate) while tracking success/failure and reasons.
  • Persistent learning storage: Saves experiences to a local JSON file for future retrieval and continuous refinement.

Quick Start

Run /self-improve --goal "สร้างระบบ multi-agent แม่ลูกที่ทำงานได้จริง" --max-attempts 5 to iteratively plan, execute, evaluate, and record improvements based on your attempt history.

Frequently Asked Questions about self-improve

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

FAQPage Schema
How does a PDCA cycle automate self-improvement for multi-agent workflows?▼

A PDCA cycle automates self-improvement by planning steps, checking results, and acting to record lessons, turning multi-agent workflow attempts into stored history for continuous refinement.

What's the best way to stop repeating trial-and-error when building complex systems?▼

To stop repeating trial-and-error, use history-driven strategy selection that saves past successful approaches and failures to a local JSON file, adapting future goal execution based on previous learning history.

How do I run iterative goal execution with persistent learning history?▼

You run iterative goal execution by specifying a goal and maximum attempts, allowing the system to break the goal into smaller steps, track success or failure, and save experiences to local storage.

Does workflow automation with knowledge persistence require external dependencies?▼

Workflow automation with knowledge persistence requires no external dependencies, relying solely on persistent local storage of experiences in a JSON file to retrieve and reuse past successful approaches.

When should I not use a history-driven adaptation approach for goal execution?▼

You should not use history-driven adaptation for goal execution when tasks are single-step or non-iterative, as the PDCA mechanism requires repeated attempts and progress tracking to effectively build learning history.

Related Skills

Explore More Software Engineering Skills→
alex-voloshin-devalex-voloshin-dev
2

memory-init

Creates the .ai-skills-memory directory structure with subdirectories for logs, designs, and knowledge sync.

Community
Advanced
agentclashagentclash
25

agentclash-agent-harness-setup

Manage E2B-based coding agent evaluation harnesses and execution workflows.

Official
Advanced
langchain-ailangchain-ai
1.1k

Deep Agents Orchestration

Delegate tasks to subagents with TodoList planning and human approval gates.

Official
Advanced
djankiesdjankies

deploying-production-migrations

Execute Prisma migrate deploy for production migrations with rollback strategies.

Community
Advanced