> ## Documentation Index
> Fetch the complete documentation index at: https://docs.originhq.com/llms.txt
> Use this file to discover all available pages before exploring further.
# X-Ray
> Deep-dive traffic analysis — prompt summaries, term maps, and behavioral fingerprinting.
**X-Ray** is Origin's traffic analysis view. Where the Overview tells you *how much* AI activity is happening, X-Ray tells you *what it's about*. It builds a semantic map of every prompt passing through your endpoints and surfaces the terms, topics, and patterns that define your organization's AI usage.
## What X-Ray shows
X-Ray operates on the current scope (all agents, or a specific agent) and the selected date range. It provides:
* **Traffic summary** — a plain-language description of what the AI traffic in this scope is about
* **Term map** — the distinct semantic terms extracted from prompt content, with stopwords removed
* **Prompt window** — the raw prompt log for the selected scope, searchable and filterable
X-Ray operates on prompt *summaries* and extracted terms, not raw prompt text. Origin never stores full prompt content — only behavioral metadata and semantic fingerprints. This is by design to protect employee privacy.
## Step 1: Navigate to X-Ray
Click **X-RAY** in the Views section of the left panel. The center panel switches to the X-Ray view, scoped to whatever agent or organizational context is currently selected.
## Step 2: Read the traffic summary
The top of the center panel shows a plain-language description: *"What this scope's traffic is about — terms that travel together across the window's prompt summaries. Click a term to filter the log."*
Below the description you'll see:
* **Prompts** — number of prompts in the scan window
* **Sessions** — number of sessions in this window
* **Terms** — list of distinct extracted terms (deduplicated, stopworded)
## Step 3: Explore the Term Map
The **Term Map** tab shows a visual representation of the semantic terms extracted from this scope's traffic. Terms that appear frequently are larger; terms that co-occur are positioned nearby.
Click any term to filter the prompt log below to entries that contain that term.
## Step 4: Review prompt templates
The **Templates** tab shows recurring prompt patterns — prompts that appear frequently enough to be classified as a template rather than a one-off query. These often represent automated workflows, CI pipelines, or shared team scripts using AI.
## Step 5: Read the Descriptions tab
The **Descriptions** tab shows the prose summary generated for this scope. This is Origin's semantic digest of the AI traffic — useful for compliance documentation, audits, or simply understanding what your team is actually doing with AI.
Use the `>` arrow next to the Descriptions tab to access additional breakdowns like model distribution, token spend, and session length distributions.