OpenClaw System Prompt

1. Course Content

Course Overview

• 💡 OpenClaw adopts an innovative "files as prompts" architecture — the AI Agent's behavior, identity, memory, tool usage rules, etc., are all defined and loaded through multiple Markdown files in the workspace.

2. 📂 Switching the Workspace

• The OpenClaw system prompts and skills are stored in the workspace path: $HOME/.openclaw/workspace
• The factory defaults provide different workspace contents in Chinese and English. The contents have the same meaning; only the language differs. You can switch the language of the OpenClaw workspace documents as needed. The factory default documents are in Chinese.
• Chinese workspace (the prompt and skill contents are in Chinese)

cd $HOME/.openclaw
rsync -av --delete .workspace_zh/ workspace/

• English workspace (the prompt and skill contents are in English)

xxxxxxxxxx
cd $HOME/.openclaw
rsync -av --delete .workspace_en/ workspace/

• After the configuration is complete, restart the OpenClaw gateway to apply the configuration

xxxxxxxxxx
openclaw gateway restart

[!WARNING]

• The .workspace_en and .workspace_zh hidden folders under the $HOME/.openclaw/workspace path are template files in different languages; it is recommended not to modify them

3. AGENTS.md

• Function description

AGENTS.md defines the top-level rules for how an Agent operates in the workspace. 📖 Think of it as the Agent's "user manual" — telling the Agent how to start up, how to manage memory, what it may do, and what it must never do.

• Content structure

AGENTS.md mainly contains the following core sections:

[!NOTE]

⚠️ The factory AGENTS.md prompt restricts OpenClaw from modifying robot program code

4. SOUL.md

• Function description

SOUL.md is the Agent's "soul" and core 🧠. It defines the Agent's way of thinking and code of conduct when facing tasks — how to control the robot body, how to find and use skills, how to plan and execute actions, and under what circumstances it should pause or refuse to execute. SOUL.md shapes the Agent's core behavioral pattern as a "robot operator".

• How it works

SOUL.md is read by the Agent at session startup and serves as the underlying guiding principle for task execution. It does not specify concrete skill parameters; instead, it specifies what kind of thinking flow the Agent should use when encountering any task. For example, the Agent is required to consult the skill library before starting any task and must not execute directly from intuition or memory — this "check the documentation first, then act" habit is defined by SOUL.md.

5. IDENTITY.md

• Function description

IDENTITY.md defines the AI Agent's self-awareness and personality traits. It defines "who" the Agent is — its name, personality, language style, and basic identity positioning. This file gives the Agent a consistent personality when interacting with users.

• How it works

At the start of each session, the Agent reads IDENTITY.md and establishes its "self-awareness" based on it. This information affects the tone in which the Agent answers questions, how it proactively offers help, and the style in which it builds relationships with users. For example, an Agent set to be "friendly and gentle" will naturally use a warmer language style when replying to users.

6. HEARTBEAT.md

• Function description

HEARTBEAT.md is the configuration file that controls the heartbeat detection mechanism in OpenClaw ⏰. The heartbeat mechanism allows the Agent to periodically "wake up" to perform some background checks and maintenance tasks even when no user has proactively started a session. This is similar to a robot performing scheduled automatic inspections — the Agent can silently complete tasks such as memory organization and status checks in the background.

• How it works

The OpenClaw gateway periodically checks the contents of the HEARTBEAT.md file to decide whether to trigger the Agent's heartbeat API call:

• The file is empty or contains only comments: The heartbeat call is skipped, and the Agent will not be woken up in the background

[!NOTE]

💤 In the robot's factory configuration, the heartbeat mechanism is disabled by default.

7. 🛠️ TOOLS.md

• Function description

TOOLS.md is the Agent's localized notes on tool usage 📝. In the OpenClaw system, "Skills" define the general way tools work, while TOOLS.md records the Agent's own specific configuration and personalized settings. In short: Skills are the instruction manual (general), and TOOLS.md is your personal notes (unique to you).

• How it works

During a session, the Agent reads the relevant Skills documents to understand the general usage of tools, and also reads TOOLS.md to obtain configuration specific to its own deployment environment.

8. 👤 USER.md

• Function description

USER.md is the dedicated file the Agent uses to understand and remember user information. If IDENTITY.md is the window through which the Agent understands "who I am", then USER.md is the window through which the Agent understands "who you are dealing with". It records contextual information such as the user's basic information, preferences, and project background, helping the Agent provide more personalized and thoughtful service.

• How it works

At the start of each session, the Agent reads USER.md following the flow specified in AGENTS.md. Through this file, the Agent can:

• Know how to address the user (name, form of address)
• Understand background information such as the user's timezone and preferences
• Remember the things the user cares about and the projects in progress

9. 🧠 MEMORY.md

• Function description

MEMORY.md is the AI Agent's long-term memory file 📚. In OpenClaw's memory system, there are two levels of memory storage:

1. 📓 Daily records (memory/YYYY-MM-DD.md): Record the raw events and conversation logs that happen each day, similar to a human's "diary"
2. 🗂️ Long-term memory (MEMORY.md): The essential information distilled from the daily records, similar to a human's "long-term memory" — containing important events, lessons learned, and key decisions and insights

• How it works

The loading and use of MEMORY.md follow these rules:

• Loaded only in the main session: When a user talks directly with the Agent, the Agent reads MEMORY.md to obtain long-term memory
• Maintained autonomously by the Agent: The Agent can freely read, edit, and update MEMORY.md in the main session

[!NOTE]

In actual use, the user can:

• ✏️ Proactively write important information: Directly write content you think the Agent should remember long-term into MEMORY.md
• 👀 Observe the Agent's autonomous maintenance: After the Agent has been running for a while, view the content that has accumulated automatically in MEMORY.md
• 🔄 Manage memory content: Regularly check and organize MEMORY.md, remove outdated information, and ensure that long-term memory stays concise and accurate