AI agents call diff_revisions to retrieve information from Law Diff without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool queries Japanese e-Gov law databases and computes a textual diff (comparison) between two law texts. This is a read-only operation that retrieves historical legal data and presents it in a comparative format. There are no side effects, no data modification, no code execution, no financial impact, and no destructive operations.
From the tool's definition Tool description states it 'Fetch[es]...and return[s] a line-level diff' — retrieves and displays comparative legal text data without modification, deletion, or execution of external operations. The action is purely data retrieval and presentation.
Attacks that exploit this kind of access
Fetch two current law IDs or law revision IDs from e-Gov and return a line-level diff of comparable article/paragraph text. It is categorised as a Read tool in the Law Diff MCP Server, which means it retrieves data without modifying state.
Register the Law Diff MCP server in PolicyLayer and add a rule for diff_revisions: 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 Law Diff. Nothing to install.
diff_revisions 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 diff_revisions 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 diff_revisions. 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.
diff_revisions is provided by the Law Diff MCP server (shayouworld/law-diff-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 →