AI agents call spreader_discover_tiles to retrieve information from Cartridge without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool reads and analyzes historical iteration log data to identify repeated sequences and discover patterns. While framed as 'discovery,' the core action is querying and analyzing logs without modifying them. However, the pattern discovery could inform automated behaviors that might later be applied, hence medium severity rather than low.
From the tool's definition The tool 'spreader_discover_tiles' description states it 'Discover[s] new tile patterns from iteration logs' — an examination and pattern-matching operation on existing logs.
Attacks that exploit this kind of access
Discover new tile patterns from iteration logs — repeated sequences become candidates. It is categorised as a Read tool in the Cartridge MCP Server, which means it retrieves data without modifying state.
Register the Cartridge MCP server in PolicyLayer and add a rule for spreader_discover_tiles: 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 Cartridge. Nothing to install.
spreader_discover_tiles 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 spreader_discover_tiles 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 spreader_discover_tiles. 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.
spreader_discover_tiles is provided by the Cartridge MCP server (lucineer/cartridge-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 →