获取配置给 AI (Cursor/Windsurf) 的最佳 System Prompt，包含核心规范和工具调用指令。建议将其复制到 .cursorrules 或 .windsurfrules 文件中。
AI agents call get_prompt to retrieve information from M8 Codex MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves static configuration data (a System Prompt template) from the server and returns it for the user to copy into their local configuration files. There are no side effects, no data modifications, no code execution, and no destructive operations. It is a pure read operation that queries and returns existing documentation or configuration guidelines.
From the tool's definition Tool name is 'get_prompt' and description states it retrieves/fetches ('获取') a pre-configured System Prompt for AI editors.
Attacks that exploit this kind of access
获取配置给 AI (Cursor/Windsurf) 的最佳 System Prompt，包含核心规范和工具调用指令。建议将其复制到 .cursorrules 或 .windsurfrules 文件中。. It is categorised as a Read tool in the M8 Codex MCP MCP Server, which means it retrieves data without modifying state.
Register the M8 Codex MCP server in PolicyLayer and add a rule for get_prompt: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches M8 Codex MCP. Nothing to install.
get_prompt is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the get_prompt rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for get_prompt. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
get_prompt is provided by the M8 Codex MCP server (lipeng9401222/m8-codex-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →