feat: add unit test retry command and feedback retry UX

[jy-tan]

Summary

Add CLI support for retrying unit test runs directly, and extend unit test feedback flows so agents can attach broad run-level guidance and trigger retries from the same command.

Changes

• Add tusk unit retry plus a matching API client method for the new retry endpoint
• Extend tusk unit feedback with --retry so feedback submission can optionally trigger a new retry run in one step
• Document run_feedback.comment as broad retry guidance rather than generic feedback, and update examples to show more concrete retry instructions
• Expand tusk unit get-run next steps to explain when to retry versus when to keep changes local and make small edits
• Update help docs and the main unit README so retry guidance, command examples, and expectations stay consistent across help surfaces

[cubic-dev-ai]

1 issue found across 6 files

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="cmd/short_docs/unit/overview.md">

<violation number="1" location="cmd/short_docs/unit/overview.md:16">
P3: Fix the typo in this user-facing sentence (`may a take` → `may take`) to keep the retry guidance clear and polished.</violation>
</file>

_{Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.}

[sohil-kshirsagar]

just confirming, when agent is trying to get runs status for a commit sha, we do show multiple retries properly?