Perform semantic search across SDOF knowledge base with OpenAI embeddings
AI agents call semantic_search_sdof to retrieve information from SDOF Knowledge Base without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Semantic search is fundamentally a read-only operation that retrieves matching documents or data based on semantic similarity. It has no side effects, creates no data, executes no code, and cannot delete or modify information. The presence of sibling destructive tools (delete_*) confirms this tool does not perform deletions.
From the tool's definition Tool performs 'semantic search across SDOF knowledge base' — a retrieval operation with no modification or deletion capability. The action is querying/searching, which is characteristic of Read operations.
Attacks that exploit this kind of access
Perform semantic search across SDOF knowledge base with OpenAI embeddings. It is categorised as a Read tool in the SDOF Knowledge Base MCP Server, which means it retrieves data without modifying state.
Register the SDOF Knowledge Base MCP server in PolicyLayer and add a rule for semantic_search_sdof: 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 SDOF Knowledge Base. Nothing to install.
semantic_search_sdof 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 semantic_search_sdof 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 semantic_search_sdof. 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.
semantic_search_sdof is provided by the SDOF Knowledge Base MCP server (tgf-between-your-legs/sdof-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 →