AI agents use persist_paper to create or update resources in Library Access MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Library Access MCP environment.
This tool writes paper metadata to the local filesystem under a Research/ directory. It creates or modifies files persistently, which is a Write operation. Severity is medium because it could overwrite existing files or fill disk space, but it is generally reversible by deleting the written files.
From the tool's definition 将抓取的论文元数据固化到本地 Research/ 目录 (persists/solidifies scraped paper metadata to local Research/ directory)
Documented attack patterns abuse exactly the kind of access persist_paper gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Library Access MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for persist_paper:
{
"version": "1",
"default": "deny",
"tools": {
"persist_paper": {
"limits": [
{
"counter": "persist_paper_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}persist_paper stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
将抓取的论文元数据固化到本地 Research/ 目录. It is categorised as a Write tool in the Library Access MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Library Access MCP server in PolicyLayer and add a rule for persist_paper: 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 Library Access MCP. Nothing to install.
persist_paper 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 persist_paper 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 persist_paper. 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.
persist_paper is provided by the Library Access MCP server (yang-kun-long/library-access-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Library Access MCP, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
7 Library Access MCP tools catalogued and risk-classified — across an index of 43,000+ MCP servers.