Plugin · Claude Code · AI Agents

Agenttoolkit Altk Evolve

Agenttoolkit-altk-evolve is a Claude Code plugin for the Operate phase that helps agents self-improve through iterative learning and conversation-derived guidelines.

by AgentToolkit · github.com/AgentToolkit/altk-evolve

Run iterative kaizen-style improvement on your Claude agent using conversation-derived guidelines so behavior gets better over time.

8
GitHub stars
0
Installs
0
Community votes
One vote per signed-in builder - it helps surface the tools the community actually relies on.
Install

Add it to Claude Code

Install the plugin in Claude Code. One command, paste-ready.

Install the plugin
/plugin install agenttoolkit-altk-evolve@AgentToolkit/altk-evolve
Add to ClaudeUse the Agent APISkillselion is itself an MCP server - your agent can fetch this config directly.
Agent API

Built to be called by your agent

Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.

Retrieve this entry with skillselion.get_details("plugin:AgentToolkit/altk-evolve") and the paste-ready config with skillselion.get_install_config("plugin:AgentToolkit/altk-evolve").

About

What it does

Agenttoolkit-altk-evolve is a Claude Code plugin focused on self-improving agents through iterations, with catalog keywords pointing at kaizen, learning from conversations, and autogenerated guidelines. For a solo builder, that means treating your coding agent’s instructions and habits as something that should tighten after real usage—not only after manual prompt edits. The journey-wide scope fits because improvement loops matter during build (tooling quality), ship (review habits), and operate (ongoing iteration), even though the directory shelves it under operate and iterate as the primary home. Install when you already run Claude Code agents regularly and want a structured evolution path instead of ad hoc “remember to do X” notes. It is not a replacement for human code review, security scanning, or formal eval harnesses; it complements those by nudging agent behavior from lived sessions.

Highlights

  • Self-improving agents through iterations (AgentToolkit/altk-evolve)
  • Kaizen and learn-from-conversations themes in catalog keywords
  • Autogenerated guidelines derived from agent conversations
  • Single-plugin productivity bundle from AgentToolkit
  • Fits builders who treat agent behavior as a product to refine, not a fixed prompt

Why builders use it

Your agent keeps repeating the same mistakes because nothing systematically turns past conversations into updated working guidelines.

After install, you can run iteration-oriented improvement so agent behavior evolves from real sessions instead of static prompts alone.

At a glance

  • Type - Plugin in AI Agents.
  • Adoption - 0 installs, 8 stars, 0 votes.

FAQ

Who is agenttoolkit-altk-evolve for?

It is for developers using Claude Code who want their agents to improve iteratively from conversations rather than freezing prompts after day one.

When should I use agenttoolkit-altk-evolve?

Use it whenever you are iterating on agent quality—especially after ship and during operate—but also while tuning agent tooling early in a project.

How do I add agenttoolkit-altk-evolve to my agent?

Install the plugin from AgentToolkit/altk-evolve (1 plugin in the repo) and enable it in Claude Code following the repository’s plugin registration steps.

Discussion

Comments

Share how you use agenttoolkit-altk-evolve, gotchas, or tips for other indie builders.

No comments yet - be the first to share how you use it.

This week for builders

Five minutes, every Monday — the tools, releases and tactics for shipping solo.

unsubscribe anytime.