Load a saved skill's code.
AI agents call load_skill to retrieve information from Code Execution MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves/reads previously saved skill code from storage. It has no side effects beyond data retrieval. Although it operates in a code execution context, loading code is distinct from executing it (that function belongs to 'execute_code'). The action is read-only and reversible, placing it in the Read category with low severity.
From the tool's definition Tool name 'load_skill' and description 'Load a saved skill's code' indicate retrieval of stored skill definitions without modification or execution.
Attacks that exploit this kind of access
Load a saved skill's code. It is categorised as a Read tool in the Code Execution MCP MCP Server, which means it retrieves data without modifying state.
Register the Code Execution MCP server in PolicyLayer and add a rule for load_skill: 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 Code Execution MCP. Nothing to install.
load_skill 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 load_skill 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 load_skill. 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.
load_skill is provided by the Code Execution MCP server (marc-shade/code-execution-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 →