---
title: "Tab: Audit"
manual: "AI Foundation"
version: "main"
permalink: "https://docs.typo3.org/permalink/nitsan/ns-t3af:ask-ai-agent-tab-audit@main"
source: "AskAiAgent/TabsAudit.rst"
rendered: "2026-10-13T14:11:01+00:00"
---

# Tab: Audit {#ask-ai-agent-tab-audit}

**Path:** **AI Foundation > AI Agent > Audit**

**What it shows:** what the AI agent records, and the most recent tool
calls.

-   **What is recorded** — tool invocations (metadata) and turn correlation
    IDs are stored. The full conversation transcript and raw prompt and
    response bodies are **not** stored in the audit log.
-   **Recent agent tool invocations** — time, tool, result
    (**Success** or **Failed**), correlation ID, and a hash
    of the arguments. No payloads are stored.

**What you do here:**

1.  Use the **Correlation** value to group all tool calls of one
    request.
1.  Investigate rows with the result **Failed**.
1.  For requests, errors, and costs across all AI features, open
    [AI Usage & Logs](https://docs.typo3.org/permalink/nitsan/ns-t3af:ai-usage-and-logs@main).