Copy an extracted bundle to a local directory (e.g. your project
AI agents use deploy_bundle to create or update resources in Claude Handoff — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Claude Handoff environment.
This tool writes/copies files from an extracted bundle into a local project directory. It creates or overwrites files on the filesystem, which is a Write operation. Severity is high because it can overwrite existing project files across potentially many locations, and could introduce malicious files if the bundle source is compromised.
From the tool's definition "Copy an extracted bundle to a local directory (e.g. your project"
Attacks that exploit this kind of access
Copy an extracted bundle to a local directory (e.g. your project. It is categorised as a Write tool in the Claude Handoff MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Claude Handoff MCP server in PolicyLayer and add a rule for deploy_bundle: 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 Claude Handoff. Nothing to install.
deploy_bundle 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 deploy_bundle 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 deploy_bundle. 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.
deploy_bundle is provided by the Claude Handoff MCP server (kiwuuu/claude-handoff-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 →