AI agents call get_error_stats to retrieve information from Pigeon without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries error statistics from the browser console history. It performs no mutations, does not execute code, and has no side effects. It is a read-only operation that retrieves aggregated data about previously collected errors. The low severity reflects that misuse would only expose error information that is already captured by the debugging context.
From the tool's definition Tool name 'get_error_stats' and description 'Counts per level, total, newest/oldest timestamps' indicate the tool retrieves and aggregates statistics about errors without modifying any state. The verb 'Counts' and 'timestamps' are retrieval operations.
Attacks that exploit this kind of access
Counts per level, total, newest/oldest timestamps. It is categorised as a Read tool in the Pigeon MCP Server, which means it retrieves data without modifying state.
Register the Pigeon MCP server in PolicyLayer and add a rule for get_error_stats: 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 Pigeon. Nothing to install.
get_error_stats 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_error_stats 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_error_stats. 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_error_stats is provided by the Pigeon MCP server (pepperonas/pigeon). 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 →