AI agents call cairntir_recall to retrieve information from Cairntir 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 information from the Cairntir SQLite database without modifying, creating, or deleting any data. Semantic search is fundamentally a read operation. Blast radius if misused is minimal: an AI agent could retrieve irrelevant or sensitive historical context, but cannot alter state or trigger destructive actions. Low severity reflects that impact is informational rather than operational.
From the tool's definition Tool performs 'semantic search across stored drawers' — a retrieval and query operation with no modifications, deletions, or external side effects. The verb 'search' and context of querying a local memory system indicate data retrieval only.
Attacks that exploit this kind of access
Semantic search across stored drawers. It is categorised as a Read tool in the Cairntir MCP Server, which means it retrieves data without modifying state.
Register the Cairntir MCP server in PolicyLayer and add a rule for cairntir_recall: 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 Cairntir. Nothing to install.
cairntir_recall 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 cairntir_recall 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 cairntir_recall. 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.
cairntir_recall is provided by the Cairntir MCP server (pnmcguire480/cairntir). 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 →