From 29d7fd5a0788d5ec37967e6800b4ec8d92b5672d Mon Sep 17 00:00:00 2001 From: "claude[bot]" <41898282+claude[bot]@users.noreply.github.com> Date: Wed, 9 Sep 2026 07:01:34 +0000 Subject: [PATCH] docs(TSP-1441): add agent model breakdown section to analytics docs Documents the Model usage by agent expandable table feature (PR #17622), which shows per-agent LLM credit consumption within the Agent Model Breakdown section. Co-Authored-By: Claude Sonnet 4.6 --- enterprise/analytics.mdx | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/enterprise/analytics.mdx b/enterprise/analytics.mdx index c89cd533..f40f055f 100644 --- a/enterprise/analytics.mdx +++ b/enterprise/analytics.mdx @@ -123,6 +123,23 @@ Each metric shows percentage change from the previous period. For example, "+45. **How to use it:** Check which actions have high execution counts to see your most-used tools. Actions with 0.00% error rates are working reliably. If an action shows error rates above 0%, investigate that integration. Sort by Credits/Exec to find expensive API calls or tool executions. + + + The Agent Model Breakdown table shows how credits are distributed across LLM models. Below it, the **Model usage by agent** table lets you expand each model row to see exactly which agents drove usage on that model. + + When you expand a model row, the table shows each contributing agent's usage: + + | Metric | What It Shows | + |--------|---------------| + | **Input tokens** | Tokens sent to the model by this agent | + | **Output tokens** | Tokens returned by the model to this agent | + | **Credits** | Credits this agent spent on this model | + | **Share** | This agent's percentage of the model's total spend | + + **How to use it:** Click any model row to expand it and reveal the per-agent breakdown. Click an agent row to navigate directly to that agent. The table respects the date range and agent filters set at the top of the analytics page. + + When the API caps results (for example, showing only the top 10 agents per model), the table displays a note indicating that results are truncated. + ---