Deep architecture analysis - entry points, hot functions, call graph, git activity (RICH output ~3-5k tokens; for first-time orientation of an unknown codebase. For targeted questions, search_structures or scan_directory are cheaper first calls)
AI agents call preview_directory to retrieve information from Scantool - File Scanner MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool analyzes and retrieves metadata about codebase structure and git history without modifying data, executing code, or triggering external operations. It is purely informational/exploratory, consistent with Read category tools like 'search', 'list', and 'get'. The large token output reflects analysis verbosity, not operational impact. Severity is low as misuse cannot harm data or systems.
From the tool's definition Tool performs 'deep architecture analysis' extracting 'entry points, hot functions, call graph, git activity' with explicitly read-only output (RICH formatted ~3-5k tokens). No modification, deletion, or execution of code indicated.
Attacks that exploit this kind of access
Deep architecture analysis - entry points, hot functions, call graph, git activity (RICH output ~3-5k tokens; for first-time orientation of an unknown codebase. For targeted questions, search_structures or scan_directory are cheaper first calls). It is categorised as a Read tool in the Scantool - File Scanner MCP MCP Server, which means it retrieves data without modifying state.
Register the Scantool - File Scanner MCP server in PolicyLayer and add a rule for preview_directory: 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 Scantool - File Scanner MCP. Nothing to install.
preview_directory 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 preview_directory 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 preview_directory. 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.
preview_directory is provided by the Scantool - File Scanner MCP server (mariusei/file-scanner-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 →