AI agents call fd_count to retrieve information from Fd without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
fd_count performs a passive file enumeration and counting operation. It retrieves metadata about files matching a pattern and returns a count. This is a read-only query with no capability to modify, delete, or execute code. The tool is safe from misuse as the worst outcome of incorrect usage would be returning an incorrect count, which causes no system harm.
From the tool's definition Tool description states 'count files matching patterns' and 'FAST FILE COUNTER'. The operation is read-only—it queries and counts existing files without modifying, deleting, or executing code on them. No side effects or irreversible operations are possible.
Attacks that exploit this kind of access
📊 FAST FILE COUNTER: Quickly count files matching patterns (faster than find | wc -l). It is categorised as a Read tool in the Fd MCP Server, which means it retrieves data without modifying state.
Register the Fd MCP server in PolicyLayer and add a rule for fd_count: 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 Fd. Nothing to install.
fd_count 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 fd_count 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 fd_count. 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.
fd_count is provided by the Fd MCP server (thhart/fd-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 →