[BUG] Routine detail page renders blank (app shell loads, content area empty)

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

Navigating to a scheduled routine's detail page (https://claude.ai/code/routines/{trigger_id}) loads the app shell (the "Code" header renders) but the content area underneath is completely blank — no run history, no output, no error message. Reproducible across multiple routines and multiple runs.

What Should Happen?

The routine detail page should show the routine's run history and the output/result of the completed run.

Error Messages/Logs

No error message or console output is shown anywhere — the page just renders blank with no indication of failure.

Steps to Reproduce

Step 1: Create a scheduled routine via the /v1/code/triggers API (POST create), with a "ccr" job config (model claude-sonnet-5, WebSearch/WebFetch tools, a single user-turn prompt event).
Step 2: Trigger a run via POST /v1/code/triggers/{id}/run.
Step 3: Wait for the run to complete (a few minutes).
Step 4: Open https://claude.ai/code/routines/{trigger_id} in a browser.

Already tried:

• Routine 1 (trig_01ErbYhMgCNiA1aCvuH8cL9U) originally created without explicitly setting persist_session (defaulted to false). Ran once (session_id: cse_014GH42J6hjGMWF1bfkmiLNP) — blank page.
• - Updated same routine to persist_session: true via partial update, confirmed via GET that persistent_session_id was populated. Ran again (session_id: cse_01SVBc2kYCsM9oZ6JNwosdn1) — still blank.
• - - Created a brand-new routine from scratch (trig_015hunf9dWsQyMy1m4bkEkSd) with persist_session: true set at creation time. Ran it (session_id: cse_01E7juGTsw6xrTUudyzL3fbc) — also blank.
• - - - In all cases, GET /v1/code/triggers/{id} shows normal metadata: enabled: true, ended_reason empty, suspension_reason empty, no errors surfaced via the API.

Environment: environment_id env_01EVqvLnki2gWaQvrByNdrxm (Default, anthropic_cloud). Routines created via the API, not the web UI. Cron schedule: 0 6 1,4 (Mon/Thu 6am UTC).

Impact: can't verify or review scheduled routine output through the web UI at all — the routine itself appears to run (metadata updates, last_fired_at changes correctly), but there's no way to see what it actually produced.

Claude Model

Sonnet (default)

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

N/A — this bug is in the claude.ai/code web app (Routines feature), not the CLI

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

This was reported and reproduced via the /v1/code/triggers HTTP API (Claude Code's remote-trigger integration), not the CLI directly — Platform/OS/Terminal fields above reflect the machine used to call the API, not where the bug manifests (which is purely in the claude.ai/code web UI).