agent-learnings-agent

Capture and classify session learnings into LRN or KNW entries in .agentic/learnings.md.

12|1|Updated Apr 7, 2026
One-click install
npx skills add https://github.com/Space-Dinosaurs/DinoStack --skill agent-learnings-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-learnings-agent
Source: https://github.com/Space-Dinosaurs/DinoStack/tree/main/.openclaw/skills/agent-learnings-agent
Command: npx skills add https://github.com/Space-Dinosaurs/DinoStack --skill agent-learnings-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Session-scoped learnings capture for AI-assisted workflows. This agent records learning events and classifies them as LRN (bug-fix) or KNW (knowledge), writing structured entries to .agentic/learnings.md and optionally to MEMORY.md, enabling knowledge reuse across tasks while enforcing dedup and soft-fail semantics.

Core Features & Use Cases

  • Spawns once per session and remains active, listening for mandatory trigger events and producing learnings.
  • Writes LRN/KNW entries using a canonical schema and optionally updates MEMORY.md for project-affecting facts.
  • Deduplicates similar entries, caps total learnings per session, and soft-fails on errors to keep the conductor responsive.

Quick Start

Trigger a learning event to start capturing a session's learnings and writing entries to .agentic/learnings.md.

Frequently Asked Questions about agent-learnings-agent

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

FAQPage Schema
How do I capture session learnings for AI-assisted workflows?

Session learnings are captured by triggering a learning event, which spawns a background agent that classifies entries as LRN or KNW and appends them to .agentic/learnings.md for reuse.

How does deduplication work for session memory management?

Deduplication for session memory management works by comparing new learning entries against existing ones in .agentic/learnings.md, preventing redundant records while enforcing a cap on total learnings per session.

What is the difference between LRN and KNW entries in learnings.md?

LRN entries capture bug-fix learnings while KNW entries record general knowledge, both formatted using a canonical schema to structure project-affecting facts for AI session reuse.

Can I automatically update MEMORY.md with project-affecting facts from a session?

Yes, the agent optionally updates MEMORY.md with project-affecting facts classified during the session, writing structured entries alongside the primary .agentic/learnings.md file.

What happens when the learning agent encounters errors during session management?

When the learning agent encounters errors during session management, it uses soft-fail semantics to keep the conductor responsive, preventing capture failures from blocking the active workflow.

Do I need conductor events to start logging learnings in the background?

Yes, conductor events are required as the agent reads mandatory trigger events to spawn once per session, listening for learnings and writing structured entries to the learnings.md file.