GitHub Repository

The self-improving sandboxed and open-source AI agent. With persistent memory and scheduling.

9 starsPython

A self upgrading agent that learns from failure

Name: A self upgrading agent that learns from failure
Availability: InStock
Author: grimm76321

by grimm76321·May 5, 2026·1 point·0 comments

Visit Project View on HN

AI Analysis

●●SolidWizardryBold Bet

Agent writes its own Python tools and saves rules to avoid repeating mistakes.

Strengths

•Zero-token monitoring via dynamically written local Python scripts.
•Sandboxed execution in Docker with domain allowlists and approval gates.
•Persistent RAG memory using TF-IDF semantic search for past experiences.

Weaknesses

•Keyword-based rule retrieval is less robust than vector-based adaptation.
•Complex setup requiring Docker and specific environment configuration.

Similar Projects

AI/ML●●Solid

Self-improving sandboxed agent with memory and scheduling

Sandboxed agent that writes its own Python tools and remembers mistakes in JSON.

Ship ItNiche Gem

grimm8000

222mo ago

AI/ML●●Solid

Sandboxed agent that improves its tools, prompts and adapts to failure

Writes errors to adaptation.json so it never makes the same mistake twice.

WizardryBold Bet

grimm8000

212mo ago

AI/ML●●Solid

AgenC – an agentic work factory focused on self-upgrading

Someone actually solved the annoying UX around running dozens of agent sessions: per-agent sandboxing, a command palette with hotkeys, and built-in secrets injection make spinning up and triaging failed runs painless. There’s also an on-board assistant that can configure and drive the factory — smart move for reducing CLI friction. The tradeoff is obvious: it’s tightly coupled to macOS and Claude, so it’s brilliant if you live in that ecosystem but niche otherwise.

Rabbit HoleNiche Gem

mieubrisse

104mo ago

Developer Tools●●Solid