AI agents invoke git_commit_push to trigger actions in Portfolio. 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 executes Git commit and push operations against a remote repository. While it has a dry-run guard, once confirmed it pushes code changes externally and cannot be trivially undone (pushed commits require a force-push or revert). It spans Write and Execute; Execute wins because it triggers external operations (remote push) whose effects depend on arguments and the current repo state.
From the tool's definition 'Commit and push changes' — triggers external Git operations (commit + push to remote); 'set confirmed:true to execute' confirms it performs real external side effects
Attacks that exploit this kind of access
Commit and push changes. First call returns dry-run, set confirmed:true to execute. It is categorised as a Execute tool in the Portfolio MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Portfolio MCP server in PolicyLayer and add a rule for git_commit_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 Portfolio. Nothing to install.
git_commit_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_commit_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_commit_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_commit_push is provided by the Portfolio MCP server (sohumsuthar/portfolio-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 →