Skip to content
TakoAPI
Back to skills

jarvis-memory-architecture

by psychotechv4

Coding Agents & IDEs
1.8kDownloads reported by clawskills.sh
7Stars
0Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

Universal memory architecture for AI agents.

Description

What This Skill Does

A file-based memory architecture for AI agents that persists identity and context across sessions. Covers long-term memory, daily logs, diary, cron inbox for cross-session messaging, social post tracking, and strategy notes that evolve over time.

File-based storage requires no external database or memory service, making the architecture portable across any agent runtime or infrastructure.

When to Use It

  • Preventing duplicate posts across social platforms
  • Tracking lessons learned between agent sessions
  • Passing cron job results back to the main agent session
  • Maintaining consistent agent persona across restarts
  • Building evolving strategy playbooks from repeated experience

Example Workflow

INPUT

User asks: Did the chess cron job win any games while I was away?

AGENT
  1. 1

    Reads memory/cron-inbox.md to check messages from isolated sessions

  2. 2

    Finds a chess win entry written by the cron job

  3. 3

    Writes the win to today's daily log in memory/YYYY-MM-DD.md

  4. 4

    Updates ELO and notes in MEMORY.md under the relevant section

  5. 5

    Clears processed entries from cron-inbox.md, keeping the header

OUTPUT

Reports the chess win and updated ELO, confirms inbox processed and long-term memory updated

Installation

clawhub install psychotechv4/jarvis-memory-architecture

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links