AI agents call analyze_query_error to retrieve information from Dune without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs analysis and diagnosis of failed queries—it retrieves error details and suggests corrections. It does not create, modify, delete, or execute any queries; it only reads and interprets existing query failure information. This is a passive diagnostic tool with no side effects on data or systems.
From the tool's definition Tool name is 'analyze_query_error' and description states it 'Analyze[s] failed queries and suggest[s] fixes'.
Attacks that exploit this kind of access
Analyze failed queries and suggest fixes (e.g. for 'Column not found'). It is categorised as a Read tool in the Dune MCP Server, which means it retrieves data without modifying state.
Register the Dune MCP server in PolicyLayer and add a rule for analyze_query_error: 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 Dune. Nothing to install.
analyze_query_error 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 analyze_query_error 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 analyze_query_error. 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.
analyze_query_error is provided by the Dune MCP server (nice-bills/dune-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
analyze_query_error is one line of Dune's registry record.
The record carries the whole server: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.
Teams ship this data inside their own products. See what a licence covers →