AI agents call bigquery to retrieve information from Nasa Ads without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves multiple bibliographic records in bulk from the NASA ADS database. Fetching data for reading purposes with no modifications, deletions, or external execution qualifies as a Read category tool. The blast radius is low—an AI agent cannot cause harm beyond accessing existing academic metadata.
From the tool's definition Tool description states 'Batch-fetch records' which is a retrieval operation. The context is a literature/citation database (NASA ADS), and fetching/retrieving bibliographic records has no side effects.
Attacks that exploit this kind of access
Batch-fetch records for up to ~2000 bibcodes in one call. It is categorised as a Read tool in the Nasa Ads MCP Server, which means it retrieves data without modifying state.
Register the Nasa Ads MCP server in PolicyLayer and add a rule for bigquery: 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 Nasa Ads. Nothing to install.
bigquery 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 bigquery 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 bigquery. 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.
bigquery is provided by the Nasa Ads MCP server (mubashir1osmani/nasa-ads). 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 →