SageMaker for Openclaw

A sophisticated memory and growth operating system designed to give AI agents consistent layered memory and self-driven evolution patterns.

tenured-master-chef-607
v1.0.1
Mar 16, 2026
0
929
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install sagemaker

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install sagemaker using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is SageMaker?

SageMaker serves as a comprehensive memory architecture for AI agents, functioning as a growth operating system. It enables agents to maintain state across short, mid, and long-term horizons, ensuring that lessons learned in one session are synthesized into permanent knowledge. By implementing a structured feedback loop, SageMaker allows any agent equipped with these Openclaw Skills to perform self-calibration, reducing recurring errors and improving long-term collaboration quality.

SageMaker Use Cases

  • Persistent agent learning across multiple sessions or complex tasks.
  • Structured self-improvement based on validated strengths and failure modes.
  • High-complexity task management requiring preflight checklists and post-task reflections.
  • Cross-agent knowledge transfer via standardized memory layers.

How SageMaker Works

  1. Capture raw evidence and experiences in short-term daily logs.
  2. Synthesize mid-term conclusions from recurring short-term data through a promotion loop.
  3. Promote stable constraints and methods to long-term memory or the knowledge base based on explicit evidence and confidence levels.
  4. Calibrate the internal self-model to reflect updated strengths, weaknesses, and growth themes.
  5. Execute a dual-gate logic model to ensure daily and weekly memory updates are completed and validated before task execution.

SageMaker Setup

To install this specific skill from the Openclaw Skills collection, create the required directory structure or use the bundled PowerShell installer:

powershell -File "skills/SageMaker/scripts/install.ps1"

For advanced recovery paths, you can apply the optional heartbeat template:

powershell -File "skills/SageMaker/scripts/install.ps1" -ApplyHeartbeatTemplate

Required directory artifacts include memory/short_term/, memory/mid_term/MEMORY.md, memory/long_term/MEMORY.md, memory/knowledge.md, and memory/check_memory.json.

SageMaker Data Schema & Taxonomy

SageMaker organizes data through a hierarchical taxonomy to ensure information integrity:

File Path Purpose Update Logic
memory/short_term/YYYY-MM-DD.md Raw daily evidence Pruned/Promoted daily
memory/mid_term/MEMORY.md Reusable conclusions Promoted weekly
memory/long_term/MEMORY.md Stable collaboration constraints High-confidence only
memory/knowledge.md Transferable methods (if X then Y) Cross-scenario utility
self-model.md Performance themes and uncertainties Validated growth items
memory/check_memory.json Gate state management Binary 1/0 status

SageMaker Advanced Features

  • Dual-gate synchronization model ensuring no tasks proceed without pending memory updates.
  • Mandatory Entry Quality Contract requiring reason, evidence, and confidence metrics for all promoted memory.
  • Task-coupled preflight checklists for medium and high complexity workflows.
  • Proposal-first safety protocol for modifying core governance and identity files.
  • Automated self-model calibration based on recurring failure modes and growth themes.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*