AI agents call get_tsutatsu to retrieve information from Tax Law without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries Japanese National Tax Agency (NTA) official tax law circulars and returns data. It supports abbreviated names for lookups but performs no write, delete, execute, or financial operations. Scraping from official sources for informational purposes is a Read operation. Low severity: misuse would only retrieve potentially sensitive tax interpretations, with no ability to modify systems or data.
From the tool's definition Tool description states it 'retrieves a specific notice/circular (tsutatsu) from NTA official sources' and performs 'scraping from the NTA site'. The verb '取得する' (acquire/retrieve) indicates data retrieval with no modification.
Attacks that exploit this kind of access
国税庁の通達(基本通達・措置法通達)から特定の通達を取得する。NTAサイトからスクレイピング。略称にも対応(所基通→所得税基本通達 等)。. It is categorised as a Read tool in the Tax Law MCP Server, which means it retrieves data without modifying state.
Register the Tax Law MCP server in PolicyLayer and add a rule for get_tsutatsu: 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 Tax Law. Nothing to install.
get_tsutatsu 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 get_tsutatsu 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 get_tsutatsu. 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.
get_tsutatsu is provided by the Tax Law MCP server (xqzimgz/tax-law-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 →