검색 요청을 이름 붙여 저장한다 (project_id 별 unique name).
AI agents use save_search to create or update resources in Assetcache — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Assetcache environment.
The tool creates and stores a named search configuration, which is a Write operation—data is being created or modified reversibly. It does not retrieve assets (Read), execute arbitrary code (Execute), delete anything (Destructive), or involve money (Financial).
From the tool's definition Tool name 'save_search' and description indicate it saves/stores a named search query ("검색 요청을 이름 붙여 저장한다" = 'saves a named search request'). This is a reversible create/store operation on metadata.
Attacks that exploit this kind of access
검색 요청을 이름 붙여 저장한다 (project_id 별 unique name). It is categorised as a Write tool in the Assetcache MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Assetcache MCP server in PolicyLayer and add a rule for save_search: 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 Assetcache. Nothing to install.
save_search 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 save_search 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 save_search. 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.
save_search is provided by the Assetcache MCP server (v0o0v/assetcache-mcp). 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 →