AI agents use git_stash_apply to create or update resources in Mcp Git — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Mcp Git environment.
This tool modifies the working directory and staging area by reapplying stashed changes, fitting the Write category (creates or modifies data reversibly). It is not Read (has side effects), not Destructive (changes are reversible via reset or rebase), not Execute (doesn't run arbitrary code), and not Financial.
From the tool's definition Tool name is 'git_stash_apply' and description states it 'Applies a single stashed state without removing it'. The git stash apply command modifies the working directory by reapplying previously saved changes, which constitutes creating or modifying data.
Attacks that exploit this kind of access
Applies a single stashed state without removing it. It is categorised as a Write tool in the Mcp Git MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Mcp Git MCP server in PolicyLayer and add a rule for git_stash_apply: 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 Mcp Git. Nothing to install.
git_stash_apply 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 git_stash_apply 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 git_stash_apply. 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.
git_stash_apply is provided by the Mcp Git MCP server (mcp-git). 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 →