self-learning

Convert execution traces into actionable knowledge to improve agent performance.

26|1|Updated Jan 6, 2026
One-click install
npx skills add https://github.com/gotar/opencode-config --skill self-learning
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-learning
Source: https://github.com/gotar/opencode-config/tree/main/skills/self-learning
Command: npx skills add https://github.com/gotar/opencode-config --skill self-learning

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Self-learning enables agents to convert execution traces into actionable knowledge to continuously improve performance.

Core Features & Use Cases

  • Trace capture and storage: automate recording of task execution histories into the learning workspace.
  • Insight extraction and knowledge-building: transform traces into actionable patterns and a persistent knowledge base.
  • Adaptive behavior: apply session-local, agent-specific, and system-wide adaptations to improve future tasks.

Quick Start

Capture a trace after completing a task, run analyze_traces.py to generate insights, and apply adaptations with apply_adaptation.py to validate improvements.

Frequently Asked Questions about self-learning

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

FAQPage Schema
How do I turn execution traces into actionable knowledge for multi-agent workflows?

To turn execution traces into actionable knowledge, you capture task execution histories into a learning workspace and run analysis scripts to extract insights that improve multi-agent workflow performance.

How can agents learn from recurring failures in complex tasks?

Agents learn from recurring failures by capturing execution traces after complex tasks, extracting failure patterns, and applying session-local or system-wide adaptations to prevent similar issues.

What is the best way to automate self-learning adaptation for software agents?

The best way to automate self-learning adaptation is to record execution traces, generate insights with analysis scripts, and apply behavioral changes with safeguarding rules across orchestrator, worker, or reviewer tasks.

Do I need a specific directory structure to store execution traces for self-learning?

Yes, you need to store execution traces in the .tmp/learning directory, which serves as the learning workspace for capturing histories before running scripts to extract insights and apply adaptations.

Can I apply agent-specific adaptations without affecting the entire system?

Yes, you can apply agent-specific adaptations independently, as the self-learning mechanism supports session-local, agent-specific, and system-wide changes with safeguarding rules to isolate behavioral updates.

What are the limitations of using execution traces for multi-agent adaptation?

Limitations include relying on execution trace quality and requiring manual validation of extracted insights, as adaptations are constrained by safeguarding rules to prevent unintended system-wide behavioral changes.