Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
AgentOS SDK 3.7.0 introduces required memory categorization and tagging. - Every memory must now use one of 8 standard categories with color labels, path prefixes, and tags. - Documentation adds a table of memory categories, path structures, and strict tagging rules. - Example code updated to demonstrate compliant storage and retrieval. - No API changes; this update enforces documentation and usage discipline for better organization and dashboard clarity.
What This Skill Does
AgentOS is a persistent memory and communication infrastructure for AI agents, delivered as a bash SDK. It handles storing, tagging, and semantically searching memories across sessions, along with project tracking, kanban boards, activity logging, and agent-to-agent messaging via a mesh layer.
Combines persistent memory, semantic search, project management, and inter-agent messaging in one SDK, removing the need to wire up separate tools for each concern.
When to use it
- Storing agent learnings and mistakes between sessions
- Tracking project tasks on a per-agent kanban board
- Sending messages between two running AI agents
- Caching API responses with automatic TTL expiration
- Logging daily agent activity for later review
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: save today's learnings and check for past OAuth mistakes
- 1Sources agentos.sh and sets AGENTOS_API_KEY, AGENTOS_BASE_URL, and AGENTOS_AGENT_ID
- 2Calls aos_put with AOS_SEARCHABLE=true to store the learning under /memory/learnings/2026-03-12
- 3Calls aos_search with query "OAuth mistakes" to find relevant past memories
- 4Returns results ranked by semantic similarity
Learning stored at /memory/learnings/2026-03-12 and past OAuth-related mistakes retrieved ranked by relevance
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.