describe_schema
AI agents call describe_schema to retrieve information from Postgres Safe without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Schema inspection retrieves metadata about database structure without side effects. While the description is empty (reducing confidence slightly), the tool name and context within a 'safe' PostgreSQL MCP server designed for secure queries strongly indicate this is a Read operation. Describing a schema does not create, modify, delete, or execute code—it only retrieves structural information.
From the tool's definition Tool name 'describe_schema' indicates schema inspection/retrieval with no modification capability. No description provided, but schema description is a standard read-only database operation.
Attacks that exploit this kind of access
describe_schema. It is categorised as a Read tool in the Postgres Safe MCP Server, which means it retrieves data without modifying state.
Register the Postgres Safe MCP server in PolicyLayer and add a rule for describe_schema: 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 Postgres Safe. Nothing to install.
describe_schema 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 describe_schema 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 describe_schema. 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.
describe_schema is provided by the Postgres Safe MCP server (sam-david/pg-redact-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 →