AI agents call hn_replies to retrieve information from Devfeed without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool queries and returns data about replies without side effects. It is purely informational—fetching existing data from Hacker News to surface reply notifications. This is a read operation with minimal blast radius, as misuse would only result in an AI accessing another user's HN reply history if credentials were compromised, not executing code or modifying data.
From the tool's definition Tool retrieves notifications/replies to Hacker News comments ('Check for replies to your HN comments'). No data modification, deletion, execution, or financial transaction occurs.
Attacks that exploit this kind of access
Check for replies to your HN comments. HN has no notifications — this fills the gap. It is categorised as a Read tool in the Devfeed MCP Server, which means it retrieves data without modifying state.
Register the Devfeed MCP server in PolicyLayer and add a rule for hn_replies: 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 Devfeed. Nothing to install.
hn_replies 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 hn_replies 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 hn_replies. 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.
hn_replies is provided by the Devfeed MCP server (upgpt-ai/devfeed-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 →