feat: add web_test, build_runner, and enhanced managed test

[tsavo-at-pieces]

Summary

• Add web_test CI feature: standalone Ubuntu job with deterministic Chrome provisioning via browser-actions/setup-chrome@v2, configurable concurrency and optional test path filtering
• Enable build_runner feature: runs dart run build_runner build --delete-conflicting-outputs before analyze/test steps to regenerate .g.dart codegen files
• Enhance managed test: pipe output through tee for log capture, use PIPESTATUS for correct exit codes, upload test logs as artifacts
• Add web_test validation (concurrency bounds, path traversal/injection safety) with 16 new tests
• Document both features in SETUP.md and USAGE.md

Test plan

• All 88+ existing + new validation tests pass (dart test)
• Generate CI with web_test: true — verify web-test job with Chrome setup steps
• Generate CI with web_test: false — verify no web-test content
• Verify build_runner steps appear in analyze and test jobs
• Verify managed test log capture and artifact upload in generated workflow

🤖 Generated with Claude Code

[Copilot]

Pull request overview

Adds new CI-generation capabilities to runtime_ci_tooling, including an optional Chrome-based web test job, optional build_runner pre-steps, and improved managed test logging/artifacts for better debuggability in GitHub Actions.

Changes:

• Add ci.features.web_test and ci.web_test (concurrency + path filtering) with validation tests.
• Add optional build_runner steps to generated workflows.
• Enhance managed test execution to capture logs via file reporters + shell tee, and upload logs as artifacts.

Reviewed changes

Copilot reviewed 11 out of 11 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
lib/src/cli/utils/workflow_generator.dart	Adds web_test feature flag + web_test config resolution and validation.
templates/github/workflows/ci.skeleton.yaml	Updates managed test steps (tee + artifacts) and introduces a new web-test job.
templates/config.json	Documents and adds default config for web_test and feature flag.
test/workflow_generator_test.dart	Adds validation test coverage for ci.web_test config.
lib/src/cli/manage_cicd.dart	Enhances managed test runner with structured reporters, log files, and job summary.
lib/src/cli/commands/test_command.dart	Adds similar output capture + summary behavior for CLI test command.
USAGE.md	Documents new optional build_runner / web_test features.
SETUP.md	Adds configuration table entries for build_runner and web_test.
.runtime_ci/template_versions.json	Updates template hashes/timestamps for the changed templates.
.runtime_ci/config.json	Enables managed_test + build_runner for this repo’s generated workflows.
.github/workflows/ci.yaml	Regenerated workflow showing build_runner + managed test log capture/upload.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[cursor]

PR Summary

Medium Risk
Touches CI workflow generation and test execution/reporting, which can affect reliability and signal in all pipelines. New config validation reduces injection risk, but changes may alter test behavior/output and artifact retention across platforms.

Overview
Adds new CI capabilities driven by config: optional build_runner execution before analyze/test and a new optional web-test job that provisions Chrome on Ubuntu and runs dart test -p chrome with configurable concurrency and validated path filtering.

Upgrades managed test execution to capture full stdout/stderr and structured JSON results, write rich $GITHUB_STEP_SUMMARY output, and always upload test logs as artifacts; generated workflows now run tests via bash with pipefail and tee to preserve correct exit codes.

Updates templates/default config and docs to expose ci.features.web_test, ci.web_test settings, and enables managed_test/build_runner in this repo’s CI; adds extensive validation + render tests to prevent unsafe web_test.paths interpolation and enforce concurrency bounds.

^{Written by Cursor Bugbot for commit 0b65313. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 2 potential issues.

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

Comment @cursor review or bugbot run to trigger another review on this PR

[tsavo-at-pieces]

Final Verification Against HEAD

I re-verified all 21 adversarial findings against current HEAD on feat/enhanced-managed-test.

Verdict

All 21 findings are resolved in current HEAD (fixed in code or invalidated by current implementation).

Per-issue status (1–21)

1. Byte vs char size guard — fixed (utf8.encode(...) byte accounting in step summary writes).
2. Invalid UTF-8 decode crash risk — fixed (Utf8Decoder(allowMalformed: true) in test stream handling).
3. Unbounded test output buffers — fixed (bounded buffers for test output and pub get).
4. Skeleton/rendered PIPESTATUS drift — resolved (no PIPESTATUS usage; set -o pipefail path is consistent).
5. SIGTERM-only timeout kill risk — fixed (kill escalation strategy in process timeout handling).
6. Stream subscription cleanup path — fixed (cleanup occurs in finally).
7. Missing single-platform PLATFORM_ID — fixed (single-platform ID is generated and used).
8. pub get timeout robustness — fixed (timeout-aware pub get execution path with termination).
9. _runTest/repoRoot contract mismatch — fixed (manage_cicd delegates with explicit repoRoot).
10. Missing unit tests for primary feature — fixed (new/expanded tests for command + summary/result utilities).
11. Artifact upload on cancelled runs (always()) — fixed (success() || failure()).
12. Exit without flush risk — fixed (exitWithCode() flushes stdout/stderr before exit).
13. Empty NDJSON marked as parsed — fixed (empty input remains unparsed).
14. Malformed JSON warning flood — fixed (circuit-breaker warning cap).
15. Sub-package tests lacking structured capture — fixed (JSON/expanded reporters + parsed summaries).
16. Weak render assertion — fixed (render tests parse YAML and assert structure).
17. collapsible() content escaping — fixed (content is escaped).
18. _codeFence pathological behavior — fixed (bounded max-run strategy).
19. Unbounded failures list growth — fixed (failures list cap + truncation behavior).
20. Retention-days inconsistency — fixed (single template variable policy applied consistently).
21. Windows pub-cache path — fixed (platform-aware cache path using LOCALAPPDATA on Windows).

Corrections to previous draft notes

• #16 is not N/A anymore; it is now explicitly covered by stronger workflow render tests.
• “Single-platform web-test independence” is outdated wording; current intended behavior is dependency on analyze-and-test in single-platform mode.
• Test count in previous note is stale; current full suite is 257 passing tests.

Current validation snapshot

• dart analyze ✅
• dart test ✅ (257 passed)