AI agents use fabric_git_create_pull_request to create or update resources in Git Steer — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Git Steer environment.
Pull request creation is a Write operation because it creates new data structures in the repository that can be closed/deleted if needed. However, severity is elevated to 'high' rather than 'medium' because in an autonomous agent context, maliciously created pull requests could: (1) inject malicious code into review workflows, (2) trigger CI/CD pipelines with side effects, (3) pollute repository history, or (4)…
From the tool's definition Tool description states 'Open a pull request' which creates new data (a PR object) in a reversible manner.
Attacks that exploit this kind of access
[git-fabric] Open a pull request. It is categorised as a Write tool in the Git Steer MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Git Steer MCP server in PolicyLayer and add a rule for fabric_git_create_pull_request: 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 Steer. Nothing to install.
fabric_git_create_pull_request is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the fabric_git_create_pull_request 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 fabric_git_create_pull_request. 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.
fabric_git_create_pull_request is provided by the Git Steer MCP server (ry-ops/git-steer). 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 →