Handle failed hosted eval creation responses

[d42me]

Summary

• treat hosted eval create responses with status FAILED as errors even when an evaluation_id is returned
• surface the backend error to users instead of printing a false success message
• add unit coverage for helper-level and CLI-level failure handling

Testing

• pytest packages/prime/tests/test_hosted_eval.py -q

---

Note

Medium Risk
Changes hosted eval creation/launch flow to treat status=FAILED responses as errors and exit non-zero, which can alter CLI behavior for users and CI scripts. Risk is limited to CLI error-handling paths and is covered by new unit tests.

Overview
Hosted eval creation failures are now surfaced correctly in the CLI. When the hosted-eval create API returns status=FAILED, the command collects the returned evaluation_id/evaluation_ids, prints the backend error, provides viewer/log commands, and exits with code 1 instead of showing a false success.

Adds helper logic to normalize singular vs plural evaluation IDs and expands tests to cover failed create responses (including plural IDs and empty evaluation_ids fallback).

^{Written by Cursor Bugbot for commit 05651b0. This will update automatically on new commits. Configure here.}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 781c7cde6b

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}