skills.rest
sakamoto-family-smile
sakamoto-family-smileCOMMUNITY
Explore all skills by this author →
GitHub

agentic-os

Coordinate persistent multi-agent workflows in Claude Code with file-based memory.

Updated Apr 13, 2026
Search Tags:#automation#multi-agent#orchestration#memory#kernel#claude-code#persistent
One-click install
npx skills add https://github.com/sakamoto-family-smile/agent_monorepo --skill agentic-os
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agentic-os
Source: https://github.com/sakamoto-family-smile/agent_monorepo/tree/main/.claude/skills/ecc/agentic-os
Command: npx skills add https://github.com/sakamoto-family-smile/agent_monorepo --skill agentic-os

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Building reliable, persistent multi-agent workflows inside Claude Code by organizing a kernel, specialist agents, slash commands, and a file-based memory layer without external databases.

Core Features & Use Cases

  • Kernel-based orchestration: a central CLAUDE.md kernel coordinates routing, identity, and policies for agents.
  • Layered design: agents/, .claude/commands, scripts/, data/ provide modular, auditable structure for long-running tasks.
  • Persistent memory: file-based state that survives sessions and restarts for continuous context and decision records.

Quick Start

Create a project root with CLAUDE.md as the kernel and populate agents/, scripts/, references/, assets/, and data/ to start building your persistent multi-agent OS.

Frequently Asked Questions about agentic-os

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

FAQPage Schema
How do I build persistent multi-agent workflows in Claude Code?▼

Build persistent multi-agent workflows in Claude Code by setting up a central CLAUDE.md kernel to coordinate routing and policies, using modular agents, slash commands, and file-based memory for state that survives restarts.

Can I orchestrate multiple agents in Claude Code without external databases?▼

Yes, you can orchestrate multiple agents without external databases by using a file-based memory layer within the data/ directory to store continuous context and decision records across sessions.

What is kernel-based orchestration for multi-agent systems?▼

Kernel-based orchestration uses a central CLAUDE.md file to coordinate routing, identity, and policies for specialist agents, providing an auditable structure for long-running tasks.

How do I set up a multi-agent OS project structure?▼

Set up a multi-agent OS by creating a project root with CLAUDE.md as the kernel, then populate agents/, .claude/commands, scripts/, references/, assets/, and data/ directories to start orchestrating tasks.

Does file-based memory survive Claude Code session restarts?▼

File-based memory survives session restarts by storing continuous context and decision records locally, ensuring persistent state for long-running projects without relying on external infrastructure.

What are the limitations of using a file-based memory layer for agent orchestration?▼

File-based memory layers lack concurrent transactional guarantees compared to external databases, meaning high-concurrency agent writes may experience race conditions if not sequentially scheduled.

Related Skills

Explore More Software Engineering Skills→
Kgan01Kgan01
6

project-scaffold

Generates hierarchical MDOT-compliant documentation for your project's needs.

Community
Intermediate
vrcmsvrcms
1

autonomous-agent-harness

Orchestrate scheduled autonomous agent workflows with persistent memory and MCP tool execution.

Community
Advanced
kurohit2kurohit2
1

Command Development

Automate structured Claude Code slash command development with YAML frontmatter and namespaces.

Community
Advanced
dejavoo21dejavoo21

hive-mind-advanced

Coordinate multi-agent systems with queen-led orchestration and configurable consensus algorithms.

Community
Advanced