Description
Experienced LLMs is an innovative open-source tool designed to enhance AI coding agents through self-education. Developed by Andrey Eremin, this engine allows AI agents to learn from daily feedback and mistake corrections, making them more tailored to individual user needs over time. Unlike traditional AI memory, Experienced LLMs focuses on self-education, capturing heuristics offline in SQLite and compiling them into clean operating directives.
Read More
The tool operates through a day-to-day learning cycle, capturing feedback during coding sessions and consolidating it overnight. This process involves logging corrections and edge cases locally, which are then reasoned over by an LLM to supersede outdated directives, deduplicate overlaps, and sort rules into clean categories. The next morning, agents start by reading these compiled directives, which are compact and free of unnecessary comments.
Designed with privacy and simplicity in mind, Experienced LLMs guarantees strict containment to specific files, ensuring zero repository scanning, indexing, or telemetry. It employs LLM-only consolidation for semantic deduplication and contradiction resolution, making it a robust tool for developers.
The tool is compatible with various platforms, including macOS, Linux, native Windows 11, and Windows 11 with WSL Hybrid environments. It is packaged as a clean open skill compatible with Antigravity, Claude Code, Cursor, and Copilot, and is available for free under the MIT license. Developers can access the repository on GitHub to inspect the code or start tailoring their local agents today.
- CategoryAI Coding Assistants
- Pricing ModelFree
- DeploymentLocal
- PlatformsmacOS, Linux, Windows 11
Experienced LLMs's Core Features
Open-source self-education engine
Captures daily feedback and corrections
Compiles heuristics into SQLite
Operates through a daily learning cycle
Ensures privacy with strict file containment
LLM-only consolidation for deduplication
Cross-platform compatibility
Free under MIT license
How to use Experienced LLMs?
Install: Set up the tool from GitHub
Configure: Adjust settings for your environment
Use: Log feedback during coding sessions
Optimize: Review and refine compiled directives
Experienced LLMs's Use Cases
- Agent Tailoring
- Feedback Integration
- Privacy-Conscious Development
- Cross-Platform AI Development
- Open-Source Collaboration
