AI agents call get_primary_keys to retrieve information from Mcp Odbc without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries table metadata (primary key information) which is a read-only operation with no side effects. It falls clearly into the Read category. Severity is low because the blast radius of misuse is minimal—an attacker can only discover schema information, which does not directly compromise data confidentiality, integrity, or availability.
From the tool's definition Tool name is 'get_primary_keys' and description states it 'Get primary key columns for a table.' This is schema introspection that retrieves metadata about table structure without modifying or executing operations on data.
Attacks that exploit this kind of access
Get primary key columns for a table. It is categorised as a Read tool in the Mcp Odbc MCP Server, which means it retrieves data without modifying state.
Register the Mcp Odbc MCP server in PolicyLayer and add a rule for get_primary_keys: 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 Mcp Odbc. Nothing to install.
get_primary_keys 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_primary_keys 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_primary_keys. 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_primary_keys is provided by the Mcp Odbc MCP server (phil-cheesman/mcp-odbc). 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 →