Self Improving Python Packages

bitranox-skills

Claude Code plugin marketplace: the bitranox skill collection (82 skills) plus a self-improve Stop hook. Skills invoke as /bitranox:<skill>.

veles-ai

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

areev

Self-improving agents, governed. Areev is the substrate for adaptive agents — agents that get better from their own history, under human authority, in steps you can inspect, undo, and re-measure.

chimera-agent

Open-source AI agent that reasons by blending many AI models, does real work on its own, and keeps learning — safe, self-hostable, and fully yours.

leeroo-kapso

Kapso by Leeroo: Long-running agents that optimize AI and Data systems, and learn from every experience. #1 open-source on MLE-Bench; ALE-Bench; RelBench

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.

langstage-hermes

A self-improving reflection / skill-creation agent on LangGraph — the LangStage reference agent

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.

sacpaint

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.

simit

SIMIT: Self-Improving Vision-Language Models via Imagination at Test-Time, official pip library of the paper, SIMIT-ICL: vision-language models that imagine their own in-context demonstrations at test time

acet

A modular research framework for agentic context engineering. It captures interaction-derived context deltas, performs semantic retrieval, ranking and curation, and re-injects validated knowledge into LLM workflows, enabling continuous adaptation without model retraining across offline and online loops.

ag2-sparrow

My AI Stand. Realtime by day, rewriting itself by night. Summon my AI superpower.

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.

self-improving-loop

Rollback-first reliability layer for AI agents: trace failures, apply guarded changes, and rollback on regression.

teleclaude

Build self-improving Telegram bots with Claude Code. Agentic self-development framework - talk to your running app, it rewrites its own code.

lorekeeper-mcp

Self-improving memory for AI agents

deepagent-hermes

A self-improving reflection / skill-creation agent on LangGraph — the LangStage reference agent

areev-crewai

Self-improving agents, governed. Areev is the substrate for adaptive agents — agents that get better from their own history, under human authority, in steps you can inspect, undo, and re-measure.

cogmem

A self-improving, verifiable memory layer for AI coding agents.

code-audit-validator

Deterministic conformance checker for AUDIT.md agent-audit outputs — CyberSkill code-audit-framework

solo-founder-os

Self-evolving 10-agent stack for one-person companies. Cron + eval + reflexion + skill library + PR-gated evolver. pip install solo-founder-os · 513 tests · MIT.

a-evolve

The official repository of "Position: Agentic Evolution is the Path to Evolving LLMs".

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

🔨 Kyoko is the all-in-one, fully local tool for debugging and improving your AI agents.