Generate or update semantic embeddings for code entities to enable semantic search
AI agents use update_embeddings to create or update resources in CodeRAG — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your CodeRAG environment.
The tool modifies data (embeddings) but does not delete, destroy, or execute arbitrary code. It is reversible—embeddings can be regenerated or overwritten without permanent loss. While the knowledge graph is important infrastructure, updating embeddings is a standard write operation.
From the tool's definition Tool name 'update_embeddings' and description 'Generate or update semantic embeddings for code entities' indicate a reversible modification operation. Creates or modifies embedding data stored in the Neo4J knowledge graph.
Documented attack patterns abuse exactly the kind of access update_embeddings gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and CodeRAG, and nothing reaches the server without passing your rules. This is the rule we recommend for update_embeddings:
{
"version": "1",
"default": "deny",
"tools": {
"update_embeddings": {
"limits": [
{
"counter": "update_embeddings_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}update_embeddings stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Generate or update semantic embeddings for code entities to enable semantic search. It is categorised as a Write tool in the CodeRAG MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the CodeRAG MCP server in PolicyLayer and add a rule for update_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 CodeRAG. Nothing to install.
update_embeddings is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the update_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 update_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.
update_embeddings is provided by the CodeRAG MCP server (jonnoc/coderag). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from CodeRAG, 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.
38 CodeRAG tools catalogued and risk-classified — across an index of 43,000+ MCP servers.