AI agents use publish_app to create or update resources in Run402 — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Run402 environment.
The tool modifies project state by publishing it and adjusting visibility/discoverability settings. These are reversible changes (unpublish, change tags), so it falls under Write rather than Destructive. Severity is medium because publishing a project could expose it unintentionally or modify its discoverability in ways that affect users, but the operation is not destructive and has no direct financial impact.
From the tool's definition 'Publish a project as a forkable app. Set visibility and tags for discoverability.' — creates a publishable artifact and modifies project metadata (visibility, tags).
Attacks that exploit this kind of access
Publish a project as a forkable app. Set visibility and tags for discoverability. It is categorised as a Write tool in the Run402 MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Run402 MCP server in PolicyLayer and add a rule for publish_app: 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 Run402. Nothing to install.
publish_app 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 publish_app 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 publish_app. 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.
publish_app is provided by the Run402 MCP server (kychee-com/run402). 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 →