Clear specified caches or all caches if no names provided.
AI agents call clear_caches to permanently remove resources in Graph Rag Obsidian — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
Clearing caches removes stored data (likely ChromaDB vector embeddings and graph relationship metadata) that may be expensive or impossible to reconstruct without re-indexing the entire Obsidian vault. This is an irreversible destructive action. The blast radius is high because clearing all caches ('if no names provided') would wipe the entire RAG system's indexed knowledge base, requiring full re-ingestion.
From the tool's definition 'Clear specified caches or all caches if no names provided' — clearing caches is an irreversible purge of cached data
Attacks that exploit this kind of access
Clear specified caches or all caches if no names provided. It is categorised as a Destructive tool in the Graph Rag Obsidian MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Graph Rag Obsidian MCP server in PolicyLayer and add a rule for clear_caches: 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 Graph Rag Obsidian. Nothing to install.
clear_caches is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the clear_caches 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 clear_caches. 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.
clear_caches is provided by the Graph Rag Obsidian MCP server (nickshffer/graph-rag-mcp-server). 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 →