self-evolve

Enumerate the current session and generate reviewable skill edits awaiting user sign-off.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/thesobercoder/skills --skill self-evolve-thesobercoder
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-evolve
Source: https://github.com/thesobercoder/skills/tree/main/self-evolve
Command: npx skills add https://github.com/thesobercoder/skills --skill self-evolve-thesobercoder

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reflects on the current session to propose durable improvements to any skills that would benefit from what was learned, ensuring the changes can be codified and reviewed.

Core Features & Use Cases

  • Analyzes invoked skills, corrections, near-misses, and detected patterns to surface actionable, durable changes.
  • Proposes edits that can be applied through the repo-skill-creator workflow, with explicit user sign-off before applying.
  • Documents the rationale and edge cases to prevent repeated mistakes across future sessions.

Quick Start

Trigger self-evolve after meaningful work to propose durable skill edits and await user sign-off.

Frequently Asked Questions about self-evolve

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

FAQPage Schema
How do I turn session learnings into durable skill edits?

To turn session learnings into durable skill edits, you trigger a reflection workflow that enumerates the current chat, detects patterns, and proposes concrete changes to existing skills for review.

How does agent self-reflection improve future workflow quality assurance?

Agent self-reflection improves workflow quality assurance by analyzing invoked skills, corrections, and near-misses to surface actionable, durable changes that prevent repeated mistakes across future sessions.

Can I apply mid-session patches to invoked skills without explicit user sign-off?

No, you cannot apply mid-session patches without explicit user sign-off. Proposed edits must await user approval before applying to ensure safe modifications to invoked skills.

What is the best way to document unwritten rules detected during a chat session?

The best way to document unwritten rules is to analyze detected patterns and corrections, then codify them as concrete, reviewable changes to existing skills with documented rationale and edge cases.

Are there limitations to using reflection for mid-session patches on ongoing chats?

A key limitation is that reflection generates proposals only; it cannot autonomously modify skills. All mid-session patches and durable edits require explicit user sign-off before applying changes.