Retrieve persistent notes attached to files or symbols. Pass path to get all notes for a file. Pass query to search semantically across all annotations. Pass both to filter by file and rank by relevance.
AI agents call get_annotations to retrieve information from Local Rag without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only retrieval operation. It queries annotation data and ranks results semantically but cannot modify, create, or delete annotations (separate tools like delete_annotation handle modifications). The blast radius is minimal—worst case an agent gains visibility into existing code documentation/notes. No side effects, reversible consequences, or external operations triggered.
From the tool's definition Tool retrieves persistent notes attached to files or symbols with no modification capability. Description explicitly states 'Retrieve' and operations are query/search ('get all notes', 'search semantically', 'filter', 'rank') with no create/update/delete…
Documented attack patterns abuse exactly the kind of access get_annotations gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Local Rag, and nothing reaches the server without passing your rules. This is the rule we recommend for get_annotations:
{
"version": "1",
"default": "deny",
"tools": {
"get_annotations": {}
}
}get_annotations 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.
Retrieve persistent notes attached to files or symbols. Pass path to get all notes for a file. Pass query to search semantically across all annotations. Pass both to filter by file and rank by relevance. It is categorised as a Read tool in the Local Rag MCP Server, which means it retrieves data without modifying state.
Register the Local Rag MCP server in PolicyLayer and add a rule for get_annotations: 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 Local Rag. Nothing to install.
get_annotations 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 get_annotations 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 get_annotations. 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.
get_annotations is provided by the Local Rag MCP server (thewinci/mimirs). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Local Rag, 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.
29 Local Rag tools catalogued and risk-classified — across an index of 43,000+ MCP servers.