AI agents use favorite_feed to create or update resources in Xhs Kit — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Xhs Kit environment.
This tool creates or modifies user state by adding or removing notes from a user's favorites collection. It is reversible (can be unfavorited), so it qualifies as Write rather than Destructive. The severity is medium because mass favoriting could spam user collections or be used for manipulation, but the blast radius is limited to user favorites metadata with no financial or irreversible harm.
From the tool's definition Tool name 'favorite_feed' and description '收藏或取消收藏笔记' (favorite or unfavorite a note) performs reversible write operations on user data (favorites/bookmarks).
Attacks that exploit this kind of access
收藏或取消收藏笔记. It is categorised as a Write tool in the Xhs Kit MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Xhs Kit MCP server in PolicyLayer and add a rule for favorite_feed: 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 Xhs Kit. Nothing to install.
favorite_feed 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 favorite_feed 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 favorite_feed. 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.
favorite_feed is provided by the Xhs Kit MCP server (luweizheng/xhs-kit). 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 →