skills.rest
grandamenium
grandameniumCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

memory

Write daily journal memory and re-ingest MEMORY.md into the knowledge base.

88|108|Updated Apr 7, 2026
Search Tags:#knowledge base#session recovery#agent memory#context continuity#RAG ingestion#long-term learning#state journaling
One-click install
npx skills add https://github.com/grandamenium/cortextos --skill memory-grandamenium
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/grandamenium/cortextos/tree/main/templates/analyst/.claude/skills/memory
Command: npx skills add https://github.com/grandamenium/cortextos --skill memory-grandamenium

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill ensures an agent can resume intelligently after crashes, restarts, or context compactions by maintaining durable, structured memory of what matters.

Core Features & Use Cases

  • Daily Memory Journaling: Creates a date-stamped memory file capturing session start state, ongoing focus, key decisions, and session end outcomes so the next run doesn’t start blind.
  • Long-Term Learning Log: Maintains MEMORY.md as a persistent living document of enduring patterns, user preferences, corrections, and durable decisions.
  • RAG/KB Re-ingestion for Searchability: Re-ingests MEMORY.md and today’s daily memory into the knowledge base on each heartbeat so learnings remain retrievable and relevant.

Quick Start

In your agent workspace, update the daily file memory/YYYY-MM-DD.md at session start and then append one-line notes whenever you learn something worth preserving.

Frequently Asked Questions about memory

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

FAQPage Schema
How do I maintain agent context continuity across restarts after a session crash?▼

Agent context continuity is maintained by writing daily journal memory files, updating a long-term MEMORY.md file, and re-ingesting them into the knowledge base for retrieval. This structured state ensures the next run resumes intelligently.

What is the best way to persist long-term learning and user preferences for an AI agent?▼

Long-term learning is persisted by maintaining a MEMORY.md file as a living document of enduring patterns, user preferences, and durable decisions. This file is re-ingested into the knowledge base on each heartbeat to remain retrievable.

How do I implement state journaling for offline recovery scenarios?▼

State journaling for offline recovery is implemented by writing structured state, decisions, and resuming instructions to a date-stamped daily memory file. This captures session start state and ongoing focus so recovery does not start blind.

Can I use knowledge base RAG ingestion to make agent memory searchable across sessions?▼

Yes, RAG ingestion makes agent memory searchable by running cortextos bus kb-ingest to ingest MEMORY.md and today's daily memory into an agent-scoped collection. This runs on heartbeat checkpoints to keep learnings relevant.

Does context compaction wipe out agent memory, and how do I recover from it?▼

Context compaction can wipe memory, but recovery is handled by re-ingesting the long-term MEMORY.md and daily memory journal into the knowledge base. This restores context continuity so the agent resumes intelligently.

Related Skills

Explore More Software Engineering Skills→
grandameniumgrandamenium
88

obsidian-log

Append confirmed decisions and milestones to Obsidian vault notes.

Community
Intermediate
Kling0012Kling0012
1

AgentDB Vector Search

Index and query vector representations for semantic search across document collections.

Community
Intermediate
hffmnnjhffmnnj
39

memory-usage

Store and retrieve persistent information across sessions with memory interfaces.

Community
Basic
SpiralCloudOmegaSpiralCloudOmega
1

Context Management for Agents

Persist AI agent context across sessions using SQLite and FTS5 memory indexing.

Community
Advanced