Read attachment content. Accepts parent item key or attachment key — auto-detects type and extracts content.
AI agents call read_attachment to retrieve information from Zotero Assistant without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and reads attachment content from Zotero library items. It performs query/fetch operations only—no data is created, modified, deleted, or executed. The only risk is potential information disclosure if attachments contain sensitive data, but this is a standard read operation with low blast radius for misuse.
From the tool's definition Tool name 'read_attachment' and description 'Read attachment content' explicitly indicate data retrieval with no modification. The function accepts keys to identify and extract existing content without side effects.
Attacks that exploit this kind of access
Read attachment content. Accepts parent item key or attachment key — auto-detects type and extracts content. It is categorised as a Read tool in the Zotero Assistant MCP Server, which means it retrieves data without modifying state.
Register the Zotero Assistant MCP server in PolicyLayer and add a rule for read_attachment: 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 Zotero Assistant. Nothing to install.
read_attachment 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 read_attachment 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 read_attachment. 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.
read_attachment is provided by the Zotero Assistant MCP server (upascal/zotero-assistant-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 →