agent-memory

Name: agent-memory
Availability: InStock
Author: nima

Store agent state, architectural facts, and task outcomes in Markdown files.

25|3|Updated Jul 14, 2026

Search Tags:#context #knowledge management #memory #persistence #notes #state

Ready to run

npx skills add https://github.com/nimadorostkar/Claude-Skills-collection --skill agent-memory-nimadorostkar

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: agent-memory
Source: https://github.com/nimadorostkar/Claude-Skills-collection/tree/main/skills/agent-tooling/agent-memory
Command: npx skills add https://github.com/nimadorostkar/Claude-Skills-collection --skill agent-memory-nimadorostkar

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the issue of AI agents losing critical context, project history, or learned constraints when a conversation session ends or the context window is compacted.

Core Features & Use Cases

Persistent State Management: Enables the agent to read and write to a dedicated memory file that survives session resets.
Failure Tracking: Explicitly records failed approaches and technical constraints to prevent the agent from repeating past mistakes.
Use Case: When working on a long-term codebase migration, the agent uses this skill to store verified architectural facts and abandoned strategies, ensuring it remains consistent across multiple days of development.

Quick Start

Initialize the agent-memory skill to create a persistent state file for tracking project constraints and failed experiments.

Frequently Asked Questions about agent-memory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I maintain persistent state for an AI agent across multiple sessions?▼

To maintain persistent state across sessions, you establish a file-based memory system that stores architectural facts and historical task outcomes in structured Markdown. This prevents the loss of critical context during session resets or context window compaction.

Why does my AI agent lose project history and learned constraints during context compaction?▼

AI agents lose project history during context compaction because the context window is truncated. A persistent file-based memory system explicitly records constraints and failed approaches to a dedicated file, preventing the loss of critical information.

What is the best way to track failed approaches and technical constraints in AI workflows?▼

The best way to track failed approaches is to implement a persistent memory system that records technical constraints and abandoned strategies in a dedicated file. This prevents the agent from repeating past mistakes during long-term codebase migrations.

Can I use Markdown files to store long-term project continuity data for AI agents?▼

Yes, you can use Markdown files to store long-term project continuity data. Structured Markdown formatting ensures the agent can efficiently retrieve architectural facts and prune stale data from the persistent state file.

How do I prevent my AI agent from repeating past mistakes during a codebase migration?▼

To prevent an AI agent from repeating past mistakes during a codebase migration, initialize a persistent memory file to explicitly record failed approaches and verified architectural facts. This maintains consistency across multiple days of development.