Semantic search across stored embeddings Use when text similarity matters beyond keyword match — native Grep finds exact strings, embeddings find meaning. Pair with memory_store / agentdb_pattern-search to land the vector against your knowledge base. For literal symbol search, native Grep is faster.
AI agents call embeddings_search to retrieve information from Claude Flow without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries data (embeddings search results) without creating, modifying, or deleting anything. It has no side effects beyond returning search results. It is purely informational retrieval, making it a Read operation with low severity since misuse only risks returning irrelevant semantic matches without affecting data integrity or system state.
From the tool's definition Tool is described as 'Semantic search across stored embeddings' and explicitly compares itself to grep for 'exact strings' vs 'meaning', positioning itself as a retrieval/query tool.
Attacks that exploit this kind of access
Semantic search across stored embeddings Use when text similarity matters beyond keyword match — native Grep finds exact strings, embeddings find meaning. Pair with memory_store / agentdb_pattern-search to land the vector against your knowledge base. For literal symbol search, native Grep is faster. It is categorised as a Read tool in the Claude Flow MCP Server, which means it retrieves data without modifying state.
Register the Claude Flow MCP server in PolicyLayer and add a rule for embeddings_search: 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 Claude Flow. Nothing to install.
embeddings_search 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 embeddings_search 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 embeddings_search. 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.
embeddings_search is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.