AI agents call load_structure to retrieve information from Fairchem without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Loading a structure file is a pure read operation. It retrieves existing data from a file into memory without creating, modifying, deleting, or executing code. While the server enables subsequent Execute actions (relaxations, MD, NEB simulations), the load_structure tool itself performs no side effects and poses minimal risk if misused in isolation.
From the tool's definition Tool description states 'Load a structure from any ASE-readable file' — a retrieval operation with no modification or execution of simulations. The verb 'load' indicates data ingestion only.
Attacks that exploit this kind of access
Load a structure from any ASE-readable file (cif, xyz, traj, POSCAR, ...). It is categorised as a Read tool in the Fairchem MCP Server, which means it retrieves data without modifying state.
Register the Fairchem MCP server in PolicyLayer and add a rule for load_structure: 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 Fairchem. Nothing to install.
load_structure 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_structure 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_structure. 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_structure is provided by the Fairchem MCP server (jkitchin/fairchem-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 →