agent-self-edit is an MIT-licensed Python package for building agents that rewrite their own system prompts using feedback from execution. It is intended for developers and researchers experimenting with adaptive agent behavior and prompt optimization.
In the Frameworks & runtimes space, agent-self-edit takes a focused approach. It focuses on improving agent behavior by automatically revising system prompts based on execution feedback. agent-self-edit is an open-source project aimed at AI developers and researchers building autonomous agents. agent-self-edit is open source under the MIT license. It runs on the command line, and it can be self-hosted.
deghosal-2026 builds and maintains agent-self-edit, and it first shipped in 2026. Development happens publicly on GitHub with 73 commits in the last 90 days. Key capabilities include prompt rewriting, execution feedback, and adaptive agents.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do