AI agents call forensics to retrieve information from Kit without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool appears to read/retrieve failure datasets and replay logs from previous agent runs for analysis purposes. 'Dataset & replays' implies reading stored failure data rather than creating or modifying anything.
From the tool's definition Failure dataset & replays — close the learning loop on failed agent runs
Attacks that exploit this kind of access
Failure dataset & replays — close the learning loop on failed agent runs. It is categorised as a Read tool in the Kit MCP Server, which means it retrieves data without modifying state.
Register the Kit MCP server in PolicyLayer and add a rule for forensics: 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 Kit. Nothing to install.
forensics 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 forensics 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 forensics. 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.
forensics is provided by the Kit MCP server (luanpdd/kit-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 →