feat(runner): add transitive hot-reload for ESM via custom loader

[adelisle]

📝 Description

Adds hot-reload support to the ESM runner with auto-detection of transitive dependency changes (parity with the CJS runner).

The ESM runner now registers a dedicated loader
(src/runner-esm-loader.mjs) via module.register(). The loader tracks every resolved user file and the parent → child import relationships seen in its resolve hook, building a reverse dependency graph. When a file changes, the loader bumps a per-file version counter for that file and every transitive importer; the next import() produces a fresh ?v=<n> URL so Node re-evaluates the whole chain.

The HotReload middleware was extended to support pluggable strategies (CJS keeps the existing require.cache / module.children walker; ESM uses the loader). When the runner exposes a non-Node module graph, the middleware enumerates user files via getAllUserFiles() and, on change, resolves which services to reload via getImporters() — only the affected services are reloaded.

Other changes:

• --hot no longer clobbers a hotReload: { ... } config object (CJS and ESM runners).
• HotReload watcher callback debounced (leading edge, 100ms) to dedupe the multiple change events fs.watch fires per save on Linux.
• runner.d.ts: new ReloadStrategy interface.
• test/e2e/scenarios/hot-reload-{cjs,esm}/: new e2e scenarios validating service-file reload, 1-hop transitive dep reload, and 3-hop chain (A → B → C → service) reload for both module systems.

🎯 Relevant issues

• Should fix Hot reload doesn't work in ESM #1319

💎 Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

🚦 How Has This Been Tested?

• npm test and npm run test:e2e have been run without errors.
• HMR tests have been added to the E2E test suite.

🏁 Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have commented my code, particularly in hard-to-understand areas

This was made with the help of AI through GitHub Copilot.