AI agents call get_atlas_history to retrieve information from AGI-MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves historical data about ATLAS process execution without modifying, deleting, or executing anything. It has no side effects beyond fetching information from the system's memory/history. The blast radius of misuse is minimal—an agent could only over-query or access unintended historical records, not cause data loss or execute harmful operations.
From the tool's definition Tool name 'get_atlas_history' and description 'Retrieve ATLAS process history for a specific task' indicate a query/retrieval operation. The verb 'Retrieve' and lack of any modification, deletion, or execution language confirm this is a read-only operation.
Attacks that exploit this kind of access
Retrieve ATLAS process history for a specific task. It is categorised as a Read tool in the AGI-MCP MCP Server, which means it retrieves data without modifying state.
Register the AGI- MCP server in PolicyLayer and add a rule for get_atlas_history: 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 AGI-MCP. Nothing to install.
get_atlas_history 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_atlas_history 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_atlas_history. 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_atlas_history is provided by the AGI- MCP server (muah1987/agi-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 →