[Run Service] Ensure `WatchRunDetails` works expected

[machichima]

Command

RUN_NAME=run-1772757248 bash runs/test/scripts/watch_run_details.sh

Response

Field	Expected	Presented?	Need implement?
details	RunDetails	✅	❌
details.run_spec	task.RunSpec	❌	✅
details.action	ActionDetails	❌	✅
details.action.id	ActionIdentifier	❌	✅
details.action.metadata	ActionMetadata	❌	✅
details.action.status	ActionStatus	❌	✅
details.action.status.phase	ActionPhase	❌	✅
details.action.status.start_time	Timestamp	❌	✅
details.action.status.end_time	optional Timestamp	❌	✅ (available when the action terminated)
details.action.status.attempts	uint32	❌	✅
details.action.status.cache_status	CatalogCacheStatus	❌	✅
details.action.status.duration_ms	optional uint64	❌	✅ (available when the action terminated)
details.action.error_info	oneof result	❌	✅
details.action.abort_info	oneof result	❌	✅
details.action.spec.task	oneof spec	❌	✅
details.action.spec.trace	oneof spec	❌	✅
details.action.attempts	repeated ActionAttempt	❌	✅

[runllm]

The data sources do not provide explicit documentation or code confirming that WatchRunDetails currently returns all the fields listed in your table (such as details.action, details.action.status, details.action.spec.task, etc.). While RunDetails and ActionDetails are documented as part of the Flyte SDK, there is no evidence in the provided sources that WatchRunDetails includes all nested fields as expected in your test, nor that these are fully implemented in the response structure. Therefore, the fields marked as "Need implement? ✅" are not confirmed as present or implemented in the current SDK or API.

If you need more detail on which fields are available or how to extend the implementation, please let me know.

Sources:

• ActionDetails API reference
• Run API reference

---

This is an AI-generated response and your feedback is appreciated! Please leave a 👍 if this is helpful and 👎 if it is not.

Mention @RunLLM to ask follow-up questions. See runllm.com for more!

[0yukali0]

#take