Insert rows or update them if they conflict on a unique constraint (INSERT ... ON CONFLICT). Specify the conflict target columns and which columns to update on conflict. Uses parameterized queries.
AI agents use upsertRows to create or update resources in Postgresql — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Postgresql environment.
The tool modifies database state by inserting or updating rows based on conflict resolution. While reversible (rows can be updated or deleted later), it materially changes data. The high severity reflects that an AI agent with unrestricted access could modify critical business data across many rows. The parameterized query design mitigates injection risks but does not reduce the write impact.
From the tool's definition Upserts rows via INSERT ... ON CONFLICT, which creates or modifies data reversibly. Description explicitly states 'Insert rows or update them if they conflict on a unique constraint' and 'which columns to update on conflict'.
Attacks that exploit this kind of access
Insert rows or update them if they conflict on a unique constraint (INSERT ... ON CONFLICT). Specify the conflict target columns and which columns to update on conflict. Uses parameterized queries. It is categorised as a Write tool in the Postgresql MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Postgresql MCP server in PolicyLayer and add a rule for upsertRows: 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 Postgresql. Nothing to install.
upsertRows is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the upsertRows 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 upsertRows. 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.
upsertRows is provided by the Postgresql MCP server (vnikhilbuddhavarapu/postgresql-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 →