feat(eval): rename --evaluator to --evaluators

[jariy17]

Several commands accept multiple evaluators using the --evaluator flag. Because the flag name is singular, it's easy to mistake it for a single-value option. This pr changes it to --evaluators flag.

What changed

Renamed across all eight commands that take the list, so the CLI doesn't end up carrying two spellings for the same concept:

Command
eval batch-evaluation evaluate
eval batch-evaluation simulate
eval online-eval create
eval online-eval update
eval ondemand evaluate
eval ondemand simulate
eval batch-insights run
project add online-eval

Verification

• bun test — 3179 pass, 0 fail (224 files)
• bun run typecheck, bun run lint:check, bun run format:check — clean
• No flag("evaluator", …), flags["evaluator"], or "--evaluator" argv left anywhere in src/

Fixture tests needed no rehashing: recorded hashes cover the SDK request, not the argv, so a flag rename doesn't move them.

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)

[agentcore-devx-automation]

AgentCore Harness Review

Verdict: Looks good

Straightforward, well-scoped flag rename from --evaluator to --evaluators across the eight commands that accept a variadic list of evaluator IDs. I checked the whole tree and everything is consistent:

• All eight handlers updated: flag definition, validation error message, and flags["…"] read site.
• All corresponding test files (including fixture tests and test.each regex matchers like /--evaluator/ → /--evaluators/) updated.
• No leftover flag("evaluator", …), flags["evaluator"], or "--evaluator" argv references remain in src/.
• The two "deliberately left alone" spots check out: project remove uses "evaluator" as a resource-type enum value (src/handlers/project/remove/index.ts:49), and supportedTuiCommands("evaluator", …) in src/handlers/eval/index.tsx names a subcommand, not a flag.
• No README or docs references to the old flag needed updating.
• Since fixture hashes cover the SDK request rather than argv, no recordings need to be re-captured — as the description notes.
• No behavior change → no new telemetry surface required.

Since the refactor CLI has not shipped GA, skipping a deprecated alias is reasonable. Nothing blocking; good to merge.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 92.30769% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 97.04%. Comparing base (2a2b25f) to head (289eeae).

Files with missing lines	Patch %	Lines
.../handlers/eval/batch-evaluation/evaluate/index.tsx	75.00%	1 Missing ⚠️
src/handlers/eval/online-eval/create/index.tsx	75.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff            @@
##           refactor    #2264   +/-   ##
=========================================
  Coverage     97.04%   97.04%           
=========================================
  Files           566      566           
  Lines         39353    39353           
=========================================
  Hits          38190    38190           
  Misses         1163     1163           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)