AI agents call garden_cluster to retrieve information from Synapse without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The garden_cluster tool performs analysis and discovery operations on existing note data within an Obsidian vault. It returns information about topical groupings and suggested index notes without modifying, creating, or deleting any data. This is a read-only query operation with minimal blast radius if misused by an AI agent.
From the tool's definition Tool description states 'discover groups' and 'get a suggested landing/index note' — pure retrieval and analysis operations with no modification, deletion, or execution of external commands.
Attacks that exploit this kind of access
Use this to discover groups of topically-related notes and get a suggested landing/index note for each group. With a seed note, returns. It is categorised as a Read tool in the Synapse MCP Server, which means it retrieves data without modifying state.
Register the Synapse MCP server in PolicyLayer and add a rule for garden_cluster: 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 Synapse. Nothing to install.
garden_cluster 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 garden_cluster 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 garden_cluster. 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.
garden_cluster is provided by the Synapse MCP server (tomjrworks/synapse-obsidian). 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 →