A portable tagging protocol that defines explicit context boundaries and memory persistence rules for AI chat sessions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install context-scope-tags
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 context-scope-tags 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 Scope Tags?
Context Scope Tags provides a lightweight, portable convention for managing explicit context boundaries within AI chat environments. By using a specific set of tags at the start of a message, developers can prevent topic bleed and dictate exactly how an agent handles memory and reasoning across different tasks.
This skill is an essential addition to any library of Openclaw Skills for users who frequently switch between multiple projects or require high-precision control over an agent's memory retention. It transforms free-form conversations into structured, predictable workflows by standardizing how context is isolated, scoped, or shared globally.
Context Scope Tags Use Cases
- Switching between distinct projects in a single chat session without context leakage.
- Performing sensitive or temporary tasks that should not be stored in long-term memory.
- Explicitly defining project-specific reasoning boundaries for complex debugging.
- Permanently saving user preferences and system decisions via memory signaling tags.
How Context Scope Tags Works
- The user places one or more standardized tags, such as [ISO] or [SCOPE], at the very beginning of their chat message.
- The AI agent parses these tags as Openclaw Skills routing rules before processing the rest of the message content.
- If an isolation tag is detected, the agent clears or ignores previous session context to ensure a fresh reasoning slate.
- If a memory tag like [NOMEM] is used, the agent bypasses its long-term storage logic for that specific interaction.
- The agent generates a response that adheres to the defined boundaries, acknowledging which context was used or reused.
Context Scope Tags Setup
Integrating these Openclaw Skills into your workflow requires no complex installation, as it is a text-based protocol. You can immediately begin prefixing messages with the supported tags. For assistant-style interfaces, you can implement a help command:
# Use the shortcut to print the tag cheat sheet if your agent supports it
/context_def
Context Scope Tags Data Schema & Taxonomy
The protocol uses a prefix-based metadata taxonomy to organize how information is handled.
| Tag Category | Identifier | Behavioral Impact |
|---|---|---|
| Isolation | [ISO: topic] |
Fresh slate; ignores all prior conversation context. |
| Scoped | [SCOPE: topic] |
Restricts reasoning strictly to the named scope. |
| Global | [GLOBAL] |
Permits and identifies cross-topic context reuse. |
| Ephemeral | [NOMEM] |
Prevents the storage of the exchange in long-term memory. |
| Persistent | [REM] |
Signals that decisions should be saved to the memory backend. |
Context Scope Tags Advanced Features
- Last-tag-wins logic to resolve conflicts between overlapping tags like [REM] and [NOMEM].
- Multi-tag stacking for complex routing, such as combining isolation with memory suppression.
- Cross-platform compatibility ensuring the protocol works on Telegram, Discord, Slack, and web-based IDEs.
- Customizable slash command integration for quick access to the protocol cheat sheet.
SKILL.md
Loading
Related Openclaw Skills
Context Clean Up
A diagnostic utility that identifies prompt context bloat and generates a safe, reversible plan to reduce token costs and improve response speed.
Skill Hub
A comprehensive toolset for discovering, security vetting, and managing Openclaw Skills from multiple registries and catalogs.
OpenClaw Whisperer
A comprehensive diagnostic, error-fixing, and skill recommendation toolkit designed to streamline Openclaw Skills management and troubleshooting.
EvoLink Cheapest Image Generation
A high-performance, cost-effective tool for generating AI images directly through your AI agent using the EvoLink API.
Cron Worker Guardrails
A reliability-first checklist for hardening POSIX-based OpenClaw background workers and unattended automation against shell-specific failures.
Openclaw Agent Optimize Skill
A utility focused on refining and optimizing the performance of AI agents within the Openclaw ecosystem.
METADATA
Featured*
