AI agents call agloop_search_logs to retrieve information from Agloop without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries and retrieves log data without side effects. It is a search/filtering operation over existing logs, analogous to grep or database SELECT queries. The blast radius of misuse is minimal—an agent could potentially discover sensitive information in logs, but cannot modify, delete, or execute operations. This is a straightforward Read operation.
From the tool's definition Tool performs 'full-text search through execution logs' with case-insensitive matching across indexed fields (agent, action, task_id, phase, input/output summaries). No modification, deletion, or execution capability is described.
Attacks that exploit this kind of access
Full-text search through execution logs. Case-insensitive search across agent, action, task_id, phase, input/output summaries. It is categorised as a Read tool in the Agloop MCP Server, which means it retrieves data without modifying state.
Register the Agloop MCP server in PolicyLayer and add a rule for agloop_search_logs: 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 Agloop. Nothing to install.
agloop_search_logs 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 agloop_search_logs 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 agloop_search_logs. 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.
agloop_search_logs is provided by the Agloop MCP server (zebbern/agloop-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 →