AI agents call git_diff_unstaged to retrieve information from Mcp Git without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only operation that queries the state of the working directory and compares it to the index. It has no side effects, does not stage, commit, delete, or execute code. The blast radius if misused by an AI agent is minimal — at worst, the agent learns what changed in files, which poses no security or data integrity risk.
From the tool's definition Tool is named 'git_diff_unstaged' and described as 'Shows changes in working directory not yet staged' — it retrieves and displays file diffs without modifying any data.
Attacks that exploit this kind of access
Shows changes in working directory not yet staged. It is categorised as a Read tool in the Mcp Git MCP Server, which means it retrieves data without modifying state.
Register the Mcp Git MCP server in PolicyLayer and add a rule for git_diff_unstaged: 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_diff_unstaged 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 git_diff_unstaged 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_diff_unstaged. 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_diff_unstaged 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 →