AI agents invoke git_branch to trigger actions in Context. 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.
While listing branches is a Read operation, creating and checking out branches modifies the git repository state (creating new refs, changing HEAD). These are external operations with side effects. The most severe applicable category is Execute, since checking out a branch can change working directory state and HEAD pointer, affecting ongoing work.
From the tool's definition 'List, create, or checkout branches' — creating and checking out branches are operations that modify repository state and trigger external git operations
Attacks that exploit this kind of access
List, create, or checkout branches. It is categorised as a Execute tool in the Context MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Context MCP server in PolicyLayer and add a rule for git_branch: 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 Context. Nothing to install.
git_branch 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_branch 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_branch. 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_branch is provided by the Context MCP server (vibhasdutta/context-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 →