AI agents invoke git_push to trigger actions in LuzzyTool. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
This tool pushes Git commits to a remote repository, which is an external operation with significant side effects. While normal pushes are reversible in theory, pushing to remote branches affects shared infrastructure and can overwrite remote history (especially with force push enabled via force=true). The explicit mention of force push capability elevates severity.
From the tool's definition 推送 Git 提交到远程仓库。默认禁止推送到 main/master 和 force push,需 force=true 显式确认。
Attacks that exploit this kind of access
【推送操作】推送 Git 提交到远程仓库。默认禁止推送到 main/master 和 force push,需 force=true 显式确认。. It is categorised as a Execute tool in the LuzzyTool MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the LuzzyTool MCP server in PolicyLayer and add a rule for git_push: 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 LuzzyTool. Nothing to install.
git_push is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the git_push 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_push. 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_push is provided by the LuzzyTool MCP server (luzzymeow/luzzytool). 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 →