AI agents call project_load to retrieve information from Engineering MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool retrieves/loads a model from a git repository, which is a read-only data retrieval operation. No data is modified, deleted, or executed. While it accesses a git project, this is a standard data-fetching operation analogous to a 'get' or 'fetch' operation. The severity is low as the blast radius of loading a model is minimal—no side effects or system changes occur.
From the tool's definition Tool name 'project_load' and description 'Load a model from a git project' indicate retrieval of existing data without modification or execution of external operations.
Documented attack patterns abuse exactly the kind of access project_load gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Engineering MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for project_load:
{
"version": "1",
"default": "deny",
"tools": {
"project_load": {}
}
}project_load is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Load a model from a git project. It is categorised as a Read tool in the Engineering MCP Server MCP Server, which means it retrieves data without modifying state.
Register the Engineering MCP Server MCP server in PolicyLayer and add a rule for project_load: 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 Engineering MCP Server. Nothing to install.
project_load 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 project_load 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 project_load. 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.
project_load is provided by the Engineering MCP Server MCP server (puran-water/dexpi-sfiles-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Engineering MCP Server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
72 Engineering MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.