docs(agents): document evaluation spans in the Agents product

[ArthurKnaus]

DESCRIBE YOUR PR

Document evaluation spans (gen_ai.evaluate, such as TypeSafe Jev) on the Agents product surface.

• Add an Evaluation Spans section to the Conversations page covering the transcript row, the timeline, and the span detail (Input questions, Output answers with probability bars and confidence), with screenshots
• Add evaluations to the transcript row-type list
• Add evaluations keywords to the Agents overview

Closes TET-3059

IS YOUR CHANGE URGENT?

Help us prioritize incoming PRs by letting us know when the change needs to go live.
Select exactly one option. For deadlines, replace YYYY-MM-DD with the due date. You can update this information later by editing the PR description.

• Urgent deadline (GA date, etc.): YYYY-MM-DD
• Other deadline: YYYY-MM-DD
• No deadline: Not urgent, can wait up to 1 week+

SLA

• Teamwork makes the dream work, so please add a reviewer to your PRs.
• Please give the docs team up to 1 week to review your PR unless you've supplied a deadline.

Thanks in advance for your help!

PRE-MERGE CHECKLIST

Make sure you've checked the following before merging your changes:

• Checked Vercel preview for correctness, including links
• PR was reviewed and approved by any necessary SMEs (subject matter experts)
• PR was reviewed and approved by a member of the Sentry docs team

🤖 Generated with Claude Code

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
sentry-docs	Ready	Preview	Oct 8, 2026 1:04pm UTC

1 Skipped Deployment

Project	Deployment	Actions	Updated
develop-docs	Ignored	Preview	Oct 8, 2026 1:04pm UTC

[linear-code]

TET-3059

[cursor]

The conversations page now documents evaluation spans and adds screenshots. The plan checks client errors on that page. An issue escalates when those errors appear.

Rollout complete

Services: sentry-docs.

Plan

What changed

The conversations docs page now describes evaluation spans in the transcript, timeline, and span detail. Readers see evaluator results, question cards, and answer cards with probability bars.

This copy is live after a production deploy of sentry-docs. The URL is /product/agents/conversations/.
The Agents overview page only adds evaluations and LLM-as-judge keywords.
This change ships only from the docs/ tree, not from develop-docs/.

Risk

Bad MDX can fail the page render for every reader of that URL. Missing screenshots leave broken images in the new section. Wrong instrumentation links send readers to missing headings. This change does not alter the Agents product runtime.

Intended effect

This is a documentation change. No Sentry signal shows that the new Evaluation Spans section is live. Treat intended_effect as unknown from telemetry. absent would mean the live page still lacks that section and the three screenshots. Confirm on the live page after deploy.

Regression watch

New images and extra MDX can break render on this page. Client errors on that URL are the symptom to watch. Pageload volume for this URL was not visible, so the error count has no traffic denominator. Zero errors may mean no visits.

Signal	Baseline	Window	Comparison	Source
Client errors on /product/agents/conversations/	0 events	2026-10-07T13:00:40Z to 2026-10-08T13:00:40Z	Hold at 0. Any new event is a regression.	Sentry errors, org sentry, project docs. Query: transaction:/product/agents/conversations/ AND environment:production. Dataset errors, field count().

If errors appear, inspect events for MDX render failures or missing image requests on that page.

Not observable

• The new Evaluation Spans copy and screenshots. No content signal exists.
• Pageload and navigation volume for this URL. A traces query for transaction:"/product/agents/conversations/" in production returned no spans in this window. A second query for transactions that contain this path also returned none. Next.js serves docs from app/[[...path]]/page.tsx, so the transaction name may be the catch-all route. Traces sample at 0.3.
• LCP and CLS for this URL. A metrics query returned null percentiles.
• HTTP 404s for the three new screenshot files.
• Broken hash links to the JavaScript and Python evaluation-span instrumentation headings.

Get Slack DMs about this change: cursor.com/automations/rollouts

[cursor]

Verified in production

• The live conversations page shows the Evaluation Spans section.