Run proctor exams against unofficial mirrors to test their MCP server configurations. Proctor spins up Docker containers on Fly Machines and runs standardized exams to verify mirrors work correctly. Available exam types: - auth-check: Verifies the authentication type and whether the mirror respon...
AI agents invoke run_exam_for_mirror to trigger actions in Langfuse Observability. 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 code/operations in external infrastructure (Docker containers on Fly Machines) rather than simply reading or writing data. While the immediate purpose is testing, the underlying action is invoking compute resources and running arbitrary exam routines.
From the tool's definition Tool description states it "spins up Docker containers on Fly Machines and runs standardized exams", which involves executing external operations.
Attacks that exploit this kind of access
Run proctor exams against unofficial mirrors to test their MCP server configurations. Proctor spins up Docker containers on Fly Machines and runs standardized exams to verify mirrors work correctly. Available exam types: - auth-check: Verifies the authentication type and whether the mirror responds correctly to auth flows (e.g., OAuth2) - init-tools-list: Connects to the mirror and retrieves its list of MCP tools, verifying the server initializes properly - both: Runs both exams sequentially Mirrors without saved mcp_json configurations are automatically skipped. Results are stored server-side in a local file and a \. It is categorised as a Execute tool in the Langfuse Observability MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Langfuse Observability MCP server in PolicyLayer and add a rule for run_exam_for_mirror: 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 Langfuse Observability. Nothing to install.
run_exam_for_mirror 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 run_exam_for_mirror 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 run_exam_for_mirror. 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.
run_exam_for_mirror is provided by the Langfuse Observability MCP server (langfuse-observability-mcp-server). 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 →