Generate pgvector embeddings for memories that have none. Only useful on the Supabase backend with HUGGINGFACE_API_KEY set. Idempotent and rate-limited: call repeatedly until remaining=0. Each call costs ceil(N/32) embedding API requests.
AI agents invoke backfill_embeddings to trigger actions in mcp-Agentmemory. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
This tool executes an external embedding generation process via the HuggingFace API and writes resulting vectors to Supabase. It spans Write (storing embeddings) and Execute (triggering external API calls). The most severe applicable category is Execute due to the external API invocation with associated costs and side effects.
From the tool's definition 'Generate pgvector embeddings for memories that have none' and 'Each call costs ceil(N/32) embedding API requests' — triggers external API calls to HuggingFace and performs compute operations on the backend.
Attacks that exploit this kind of access
Generate pgvector embeddings for memories that have none. Only useful on the Supabase backend with HUGGINGFACE_API_KEY set. Idempotent and rate-limited: call repeatedly until remaining=0. Each call costs ceil(N/32) embedding API requests. It is categorised as a Execute tool in the mcp-Agentmemory MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the mcp-Agentmemory MCP server in PolicyLayer and add a rule for backfill_embeddings: 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 mcp-Agentmemory. Nothing to install.
backfill_embeddings is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the backfill_embeddings 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 backfill_embeddings. 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.
backfill_embeddings is provided by the mcp-Agentmemory MCP server (obidel/agentmemory). 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 →