AI agents call eval_factual_consistency to retrieve information from Vectara MCP server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool appears to evaluate or check factual consistency (likely comparing claims against a knowledge base), which is an analytical/assessment operation with no data modification. However, confidence is moderate because the description is empty, limiting definitive classification. The tool's placement among query and search tools on a Vectara (vector database) server reinforces Read categorization.
From the tool's definition Tool name 'eval_factual_consistency' suggests evaluation/assessment of existing data. Grouped with search_vectara, ask_vectara, and correct_hallucinations — all read-oriented operations. No side effects apparent from name.
Documented attack patterns abuse exactly the kind of access eval_factual_consistency gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Vectara MCP server, and nothing reaches the server without passing your rules. This is the rule we recommend for eval_factual_consistency:
{
"version": "1",
"default": "deny",
"tools": {
"eval_factual_consistency": {}
}
}eval_factual_consistency is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
eval_factual_consistency. It is categorised as a Read tool in the Vectara MCP server MCP Server, which means it retrieves data without modifying state.
Register the Vectara MCP server MCP server in PolicyLayer and add a rule for eval_factual_consistency: 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 Vectara MCP server. Nothing to install.
eval_factual_consistency 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 eval_factual_consistency 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 eval_factual_consistency. 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.
eval_factual_consistency is provided by the Vectara MCP server MCP server (vectara/vectara-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Vectara MCP server, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
6 Vectara MCP server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.