AI-powered automatic project context management that detects user intent and maintains context awareness.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install auto-context-manager
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install auto-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 Auto Context Manager?
Auto Context Manager is a specialized tool designed to solve the problem of context fragmentation in multi-project environments. By integrating this into your suite of Openclaw Skills, you enable your AI agent to intelligently determine which project you are discussing based on natural language triggers. It acts as a traffic controller for your sessions, ensuring that the right project-specific data and memories are prioritized.
This skill is built for developers who need to switch contexts rapidly without manual reconfiguration. It functions entirely locally, maintaining a persistent registry of projects, descriptions, and keywords to ensure that your agent always operates with the correct project-aware knowledge.
Auto Context Manager Use Cases
- Automatically switching project context during an active chat session based on message content.
- Organizing multiple codebases and ensuring Openclaw Skills access the correct repository data.
- Triggering specific memory sets based on detected project keywords like "portfolio" or "trading app."
- Managing a local directory of project metadata for quick CLI-based switching.
How Auto Context Manager Works
- The user sends a message or a specific command to the AI agent.
- The system runs the detection logic via the acm.py script to analyze the input text.
- The tool compares keywords and descriptions stored in the local projects.json file against the user's message.
- A confidence score is calculated; if a match is found, the current active project is updated.
- Subsequent actions by other Openclaw Skills utilize this active context to filter relevant information.
Auto Context Manager Setup
To get started with this project context manager, ensure you have Python installed in your environment. The configuration and project storage are handled automatically in your home directory.
# List all currently registered projects
python acm.py list
# Detect context from a specific string
python acm.py detect "I need to update my trading bot"
# Manually switch the active project context
python acm.py switch project_id
You can also integrate it directly into your Python scripts:
from auto_context_manager import AutoContextManager
acm = AutoContextManager()
acm.create_project('web-app', 'Frontend React Site', ['react', 'ui', 'frontend'], 'Main web project')
Auto Context Manager Data Schema & Taxonomy
The skill maintains a structured JSON database to track project metadata. Data is stored locally for privacy and speed.
| Component | Path / Type | Description |
|---|---|---|
| Project Config | ~/.auto-context/projects.json |
Stores project IDs, descriptions, and keywords. |
| CLI Wrapper | acm.py |
Entry point for shell-based context management. |
| Core Logic | auto_context_manager.py |
Python module handling the detection algorithms. |
| Context State | current_project |
A key in the JSON file tracking the active session context. |
Auto Context Manager Advanced Features
- Local-first architecture ensures no project metadata or keywords are sent to external APIs.
- Weighted keyword matching and confidence scores for high-accuracy project identification.
- Seamless integration with the broader ecosystem of Openclaw Skills to drive context-aware responses.
- Automated fallback to a default project to ensure session continuity even when context is ambiguous.
SKILL.md
Loading
Related Openclaw Skills
AgentMesh Governance
A zero-trust governance layer providing cryptographic identity verification, policy enforcement, and tamper-evident auditing for AI agents.
Insula Memory
A neuro-inspired internal state awareness module for AI agents to track energy, mood, and interoception.
Basal Ganglia Memory
A specialized procedural memory skill that enables AI agents to form habits and preferences through repetitive actions and reward-based learning.
Anterior Cingulate Memory
A neural-inspired conflict detection and error monitoring module that acts as a cognitive filter for AI agents.
ClawdVine
An agentic media network for generating high-quality AI videos using onchain identity and pay-per-use USDC payments.
ClawdVine
ClawdVine is an agentic media network for generating AI videos with onchain identity and pay-as-you-go USDC payments.
Featured*
