[feat][evaluation] add RunModeConfig.SkillsMode for sandbox agent skills

[bitholic]

Background

Add skills_mode to RunModeConfig, declaring how a sandbox-agent experiment combines test-case-level and experiment-level Agent Skills (merge / disable_test_case). Forward-compatible (optional field).

This is the OSS half of a two-repo change. The commercial fork (cozeloop-commercial) emits this field to the case-file experiment_info.skills_mode, and carries the test-case-level skills list to dataset_item.skills (aligned with runtime testcase.TestCase.Skills). See companion commercial MR.

Changes

• idl/thrift/.../domain/expt.thrift + domain_openapi/experiment.thrift: add skills_mode field to RunModeConfig
• backend/.../domain/entity/expt.go: RunModeConfig.SkillsMode
• backend/.../application/convertor/experiment/{expt,openapi}.go: wire the field through convertors
• backend/kitex_gen/...: regenerated via code_gen.sh

Verification

• cd backend && go build ./...

🤖 Generated with Claude Code

[CLAassistant]

All committers have signed the CLA.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

@@           Coverage Diff            @@
##             main     #645    +/-   ##
========================================
  Coverage   78.47%   78.48%            
========================================
  Files         696      696            
  Lines       85755    85978   +223     
========================================
+ Hits        67299    67478   +179     
- Misses      14520    14567    +47     
+ Partials     3936     3933     -3     

Flag	Coverage Δ
unittests	78.48% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...valuation/application/convertor/experiment/expt.go	82.65% <100.00%> (+0.05%)	⬆️
...uation/application/convertor/experiment/openapi.go	92.41% <100.00%> (-0.58%)	⬇️
backend/modules/evaluation/domain/entity/expt.go	82.24% <ø> (ø)

... and 10 files with indirect coverage changes

---

Continue to review full report in Codecov by Harness.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7469d17...a3ba7a2. Read the comment docs.

🚀 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.