memory-manager

Stores and retrieves task outcomes and agent performance across Legion workflow sessions.

1|Updated Jun 19, 2026

Search Tags:#memory management #performance tracking #agent performance #task outcomes #legion workflow

Ready to run

npx skills add https://github.com/9thLevelSoftware/legion-ascended --skill memory-manager-9thlevelsoftware

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: memory-manager
Source: https://github.com/9thLevelSoftware/legion-ascended/tree/main/skills/memory-manager
Command: npx skills add https://github.com/9thLevelSoftware/legion-ascended --skill memory-manager-9thlevelsoftware

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of storing and retrieving task outcomes and agent performance across sessions in a Legion multi-agent workflow, enabling users to reference past sessions, recall previous results, and learn from previous experiences.

Core Features & Use Cases

Cross-Session Memory: Stores and retrieves task outcomes and agent performance across sessions.
Recency and Importance Scoring: Scores recalled entries by recency and importance.
Graceful Degradation: Operates without memory files, providing fallbacks.
Use Case: Ideal for users working within a Legion multi-agent workflow who want to track and review past session outcomes, recall previous results, and learn from past experiences.

Quick Start

Run the 'memory-manager' skill to store or retrieve task outcomes and agent performance data.

Frequently Asked Questions about memory-manager

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

FAQPage Schema

How do I store and retrieve task outcomes across sessions in a Legion workflow?▼

Cross-session memory for task outcomes in Legion workflows is managed by storing structured markdown files in the `.planning/memory/` directory. The skill retrieves and scores past entries based on recency and importance.

Can I track agent performance across multiple sessions without losing previous results?▼

Yes, agent performance tracking across sessions is supported by storing structured markdown files. It retrieves previous results and scores them by recency and importance, enabling users to recall and learn from past experiences.

What is the best way to recall previous results in a multi-agent workflow?▼

Recalling previous results in a multi-agent workflow is handled by retrieving stored markdown files from the `.planning/memory/` directory. Entries are scored by recency and importance to surface the most relevant past outcomes.

Does the memory manager require existing memory files to function?▼

No, the memory manager operates with graceful degradation, meaning it functions without existing memory files by providing fallbacks. It requires access to the `.planning/memory/` directory to store new structured markdown data.

How are recalled task outcomes scored for relevance?▼

Recalled task outcomes are scored by recency and importance. This scoring mechanism ensures that the most relevant and recent agent performance data is prioritized when retrieving cross-session memory from markdown files.