Context Manager for Openclaw

An intelligent context management skill that monitors session usage and automates memory transfers to prevent data loss and ensure session continuity.

zhaog100
v1.0.0
Mar 4, 2026
0
161
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install openclaw-context-manager

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 openclaw-context-manager 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 Context Manager?

Context Manager is a sophisticated automation tool designed to solve the challenge of context window limitations in AI interactions. By operating as one of the most proactive Openclaw Skills, it monitors session tokens in real-time and executes a strategic memory transfer once usage hits a 95% threshold. This ensures that your technical insights, task progress, and critical decisions are never lost when a session ends.

The skill bridges the gap between individual sessions by synthesizing current activity into a structured memory system. It transforms the way developers interact with AI agents by removing the cognitive load of manual session tracking, allowing for a seamless transition using simple commands like /new while maintaining full project awareness.

Context Manager Use Cases

  • Managing long-duration coding sessions where context limits are frequently reached.
  • Automating the creation of daily work journals and technical decision logs.
  • Protecting active task progress by syncing state to HEARTBEAT.md before a session reset.
  • Maintaining continuity across complex multi-step project phases without manual re-contextualization.

How Context Manager Works

  1. The skill silently monitors the session_status to track the percentage of the context window used.
  2. At an 80% warning level, the skill begins background preparation of a session summary.
  3. When the 95% critical threshold is reached, the skill extracts key information including incomplete tasks and core lessons learned.
  4. It automatically updates the local memory architecture, including MEMORY.md, daily logs, and HEARTBEAT.md.
  5. A notification is sent via the configured QQ Bot to alert the user that a memory transfer has been completed.
  6. The user enters the /new command to start a fresh session, which automatically loads the recently archived context.

Context Manager Setup

To integrate this into your existing Openclaw Skills environment, ensure you are running version 2026.1.0 or higher. Use the following commands to verify your environment:

# Check OpenClaw version
openclaw --version

# Verify gateway status for notifications
openclaw gateway status

No manual configuration is required as the skill uses optimized defaults for threshold monitoring and file updates.

Context Manager Data Schema & Taxonomy

The skill organizes data into a three-level priority system to ensure logical taxonomy:

Level Target File Information Type
Level 1: Urgent MEMORY.md Current task progress, decisions, and core lessons
Level 2: Important memory/YYYY-MM-DD.md Technical insights and work journal entries
Level 3: Tasks HEARTBEAT.md Ongoing, completed, and paused task statuses

Context Manager Advanced Features

  • Real-time session parsing with sub-second detection delay for near-instant threshold triggers.
  • Intelligent session summarization that distinguishes between urgent decisions and general work logs.
  • Automated notification system through QQ Bot integration for proactive session management.
  • Zero-dependency architecture that relies on local Markdown files for maximum privacy and speed.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*