AI agents call view_starlog to retrieve information from Starlog without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool retrieves information about registry paths and file locations for starlog session data. While it doesn't execute code or modify data, the medium severity reflects that exposing raw file paths could enable an AI agent to then access sensitive session data or debug information stored in those locations.
From the tool's definition Tool description states it "returns the actual file paths to access the raw session data" - this is a retrieval operation that queries and returns information about data locations without modifying anything.
Attacks that exploit this kind of access
Get registry paths where starlog data is stored. Returns the actual file paths to access the raw session data. It is categorised as a Read tool in the Starlog MCP Server, which means it retrieves data without modifying state.
Register the Starlog MCP server in PolicyLayer and add a rule for view_starlog: 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 Starlog. Nothing to install.
view_starlog is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the view_starlog 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 view_starlog. 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.
view_starlog is provided by the Starlog MCP server (sancovp/starlog-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 →