fix(rules): surface rule failures during export

[tangcent]

Summary

Fixes #757 — AI-generated custom.* rules that throw at evaluation time no
longer silently skip every endpoint.

Problem

A throwing custom.method.is.api rule made every method evaluate to
"not an API" (runCatching{...}.getOrDefault(false) in the exporter), so
exports came back silently empty with only WARN lines in idea.log.
Three defects in one generated rule file went unnoticed because
RuleProposalValidator only did static checks and never executed the
proposed scripts:

1. it.static threw MissingPropertyException on method contexts
   (isStatic() was missing, unlike ClassContext/FieldContext).
2. it.args (property access) threw — args() is a plain method.
3. it.canonicalText() on a param context returned the element path
   (com.example.Foo#bar.userId), not the parameter's type.

Changes

• RuleFailureMonitor (new, project service) aggregates rule-evaluation
  failures per run; ExportOrchestrator brackets each run so one warning
  balloon surfaces at the end instead of a silent empty export.
• RuleEngine now logs and records throwing filters and values per
  occurrence (previously dropped silently by aggregation).
• RuleDryRunValidator (new) executes every groovy: value/filter in a
  proposal once against representative PSI contexts before staging —
  compile errors or API misses on every context kind block the proposal,
  others become reviewer notes. Wired into ProposeRuleContentTool and
  OrchestratorProposeRuleContentTool (which now use
  reviewWithDryRun).
• isStatic() added to method contexts (API symmetry with class/field
  contexts) so it.static resolves.
• Soft warning for it.canonicalText() on parameter-context keys
  (suggests type().name()).

Testing

• RuleFailureMonitorTest (6 new) — run windowing, aggregation, balloon
  content, buffer clearing.
• RuleDryRunValidatorTest (15 new) — script collection, context-kind
  selection, compile-error/API-miss classification, extra-binding-key
  skipping.
• ScriptPsiContextsTest.testMethodContext_IsStatic + RuleProposalValidatorTest
  param-canonicalText warning tests.
• Full RuleEngineTest, AntiPatternGateTest, and
  RuleProposalValidatorTest suites pass.
• ./gradlew compileTestKotlin clean.

Risks / rollback

• Dry-run validation scans up to 100 project classes and executes scripts
  at proposal time; infrastructure failures are logged and swallowed — a
  proposal is never blocked by a dry-run infra error. Keys whose
  evaluation stage needs extra bindings (api, request/response,
  collection, …) are skipped.
• Well-behaved rules behave exactly as before: no failure, no balloon.
• Failures outside a run window (e.g. background dashboard scans) are
  logged per occurrence but never ballooned.
• Revert by reverting this commit; the branch is feature-isolated.

[github-actions]

📦 Plugin has been packaged for this PR. You can download easy-api-3.1.9.252.0.zip from the GitHub Actions workflow run by clicking on the "Artifacts" dropdown.

[codecov]

Codecov Report

❌ Patch coverage is 79.68750% with 39 lines in your changes missing coverage. Please review.
✅ Project coverage is 66.927%. Comparing base (f6f7d03) to head (9186c48).

Files with missing lines	Patch %	Lines
...itangcent/easyapi/core/rule/RuleDryRunValidator.kt	80.159%	4 Missing and 21 partials ⚠️
...yapi/core/rule/context/RuleScriptContextCatalog.kt	25.000%	0 Missing and 6 partials ⚠️
...m/itangcent/easyapi/core/rule/engine/RuleEngine.kt	40.000%	2 Missing and 1 partial ⚠️
...angcent/easyapi/core/rule/RuleProposalValidator.kt	85.714%	0 Missing and 2 partials ⚠️
...ore/ai/tools/OrchestratorProposeRuleContentTool.kt	0.000%	1 Missing ⚠️
...tangcent/easyapi/core/export/ExportOrchestrator.kt	91.667%	1 Missing ⚠️
...ent/easyapi/core/rule/engine/RuleFailureMonitor.kt	95.833%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##            master      #761       +/-   ##
=============================================
+ Coverage   66.827%   66.927%   +0.100%     
=============================================
  Files          443       445        +2     
  Lines        25952     26136      +184     
  Branches      6195      6251       +56     
=============================================
+ Hits         17343     17492      +149     
- Misses        6289      6294        +5     
- Partials      2320      2350       +30     

Flag	Coverage Δ
unittests	66.927% <79.688%> (+0.100%)	⬆️

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

Files with missing lines	Coverage Δ
...nt/easyapi/core/ai/tools/ProposeRuleContentTool.kt	94.286% <100.000%> (ø)
...ent/easyapi/core/rule/context/ScriptPsiContexts.kt	49.716% <100.000%> (+0.143%)	⬆️
...ore/ai/tools/OrchestratorProposeRuleContentTool.kt	54.717% <0.000%> (ø)
...tangcent/easyapi/core/export/ExportOrchestrator.kt	66.250% <91.667%> (+3.750%)	⬆️
...ent/easyapi/core/rule/engine/RuleFailureMonitor.kt	95.833% <95.833%> (ø)
...angcent/easyapi/core/rule/RuleProposalValidator.kt	91.597% <85.714%> (-0.856%)	⬇️
...m/itangcent/easyapi/core/rule/engine/RuleEngine.kt	72.807% <40.000%> (+0.080%)	⬆️
...yapi/core/rule/context/RuleScriptContextCatalog.kt	86.059% <25.000%> (-1.338%)	⬇️
...itangcent/easyapi/core/rule/RuleDryRunValidator.kt	80.159% <80.159%> (ø)

---

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 f6f7d03...9186c48. Read the comment docs.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.