Self Improving Python Packages

veles-ai

Minimal CLI agent framework with compounding project memory — gets smarter with every session. Clean, modular, any LLM provider

threadkeeper

Multi-agent shared brain across Claude Code/Desktop, Codex, Gemini, Copilot, VS Code. Cross-session memory, self-improving skill loops, inter-agent signaling — one local MCP server.

opencastor

OpenCastor — productized open-core RCAN runtime. Gateway-as-kernel + drivers + fleet + cloud bridge + UI. Layer 4 of the OpenCastor stack. See rcan.dev/compatibility for protocol versions.

lorekeeper-mcp

Self-improving memory for AI agents

code-audit-validator

AUDIT.md — an honest, self-improving audit protocol for AI coding agents. Evidence-gated, eval-regressed, AI-agnostic. By CyberSkill.

leanlab

A self-improving lab for AI agents — evolve ML experiments against a frozen metric, or ship coding tasks through a spec → gate → review → merge loop with locked acceptance tests.

kyoko

Local-first repair loop for debugging and improving AI agents.

solo-founder-os

🧠 Self-evolving 10-agent stack for one-person companies. Cron + eval + reflexion + skill library. MIT.

acet

Agentic Context Toolkit: context delta learning for adaptive LLM agents

metabrain

A zero-dependency SQLite memory layer for AI agents that learns what works: patterns graduate to proven preferences via a built-in experiment loop.

a-evolve

Evolve any agent across any domain using any evolution algorithm — with zero human intervention

teleclaude

Telegram bot framework with Claude Code session integration

hyperflow-ai

A self-improving agent framework built on LangChain and LangGraph

self-improving-loop

A safety-first self-improvement loop for AI agents: execute → track → analyze → auto-apply → auto-rollback on regression.

taijios-soul

Archived TaijiOS legacy prototype; current main repo is yangfei222666-9/taiji.