AI agents call git_history to retrieve information from Git without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
git_history performs only read operations to inspect Git history, commits, references, and blame information. It retrieves data without modifying, creating, deleting, or executing any changes to the repository. This is a straightforward Read category tool with low severity as misuse would only expose existing information.
From the tool's definition Tool description states 'Use action=log|show|reflog|blame|lg|who to inspect commits, refs, and contributors.' All listed actions (log, show, reflog, blame, lg, who) are inspection/querying operations with no side effects.
Attacks that exploit this kind of access
History tool. Use action=log|show|reflog|blame|lg|who to inspect commits, refs, and contributors. It is categorised as a Read tool in the Git MCP Server, which means it retrieves data without modifying state.
Register the Git MCP server in PolicyLayer and add a rule for git_history: 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 Git. Nothing to install.
git_history 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_history 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_history. 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_history is provided by the Git MCP server (selfagency/git-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 →