refactor(replay): pass runtime hosts through replay and device selection

[thymikee]

Summary

Part of #3753. Part of #3747.

Pass the owning host and explicit environment through replay-port, maestro, ad-script, plugins, and device selection. Required daemon callers and fixtures are included: 68 files, 816 gross lines.

Second layer, based on #3826. #3823 remains the top gate layer. The rebase preserves the original cleanup changes and includes #3818/#3820 from main.

Validation

Head efb97886892ab3342970a2560883e34dd82d22e1.

pnpm check:affected --run --base chore/process-globals-daemon --head efb97886892ab3342970a2560883e34dd82d22e1 passed; receipt run-sfbw1g, exactCommit=true.

Coverage failures were fixture races: 7e2cae20ed replaces a directory-order assumption with explicit files; 288e6cceae waits for bridge evidence contents rather than file creation. Both fixes are inherited by all layers. Upper patches are unchanged by range-diff.

Prior full-check and Android daemon/provider evidence stays attributed to e223d37244 in #3823’s review-response comment. This round changes only test fixtures and ancestry. Full coverage and GitHub-owned lanes remain CI obligations.

[github-actions]

Size Report

Metric	Base	Current	Diff
Installed (including dependencies)	5.44 MB	5.44 MB	+769 B
Package (unpacked)	5.44 MB	5.44 MB	+769 B
Package (download)	1.64 MB	1.64 MB	+292 B

Startup median (7 runs, lower is better):

Scenario	Base	Current	Diff
CLI --version	30.6 ms	32.6 ms	+1.9 ms
CLI --help	56.2 ms	56.6 ms	+0.4 ms

[thymikee]

Coverage follow-up: 7e2cae20ed fixes the daemon progress fixture by passing explicit files; the failing assertion assumed filesystem directory order. Restacked this unchanged replay patch to 51d5a75793; range-diff is identical. Foreground check:affected passed at 51d5a75793 (run-cdxM2Z, exactCommit=true). CI is running on the pushed head.

[thymikee]

Second coverage fix: 288e6cceae polls the detached bridge’s expected cwd contents; the shell creates the evidence file before writing it, and CI read an empty file (run 39289751890). Head efb9788689 inherits both this fix and 7e2cae20ed; upper-layer range-diff is unchanged. Foreground check:affected passed (run-sfbw1g, exactCommit=true). Cold self-review found no issue in the test-only diff. CI is running on this head.

[thymikee]

The code in efb9788 looks ready: I found no problems that need a change before merge, and I did not run the tests myself. CI on this head was still pending when I looked, so Integration, Coverage and Smoke must finish green, since they exercise the replay, device-selection and daemon handler routes this PR touches. There are no conflicts. I did not review the stacked base #3826, which supplies sessionStore.host and the handler host, so #3826 must land first; I assumed both carry the same daemon host environment, based on the reads in request-router.ts. Live device evidence is not needed, because the env value is unchanged. Not blocking, and you can take or leave these: the updated tests all pass process.env as the host env, so a wrong env at a production call site would go unnoticed, and one router-level test with a distinct env (for example an AD_VAR_* entry) would pin the wiring; includePathKey in engine-flow.ts now uses dirname(flow.sourcePath) instead of process.cwd() for a sourceless runFlow, which fixes the cycle key quietly and has no test; and handlers read the host from two places (sessionStore.host.env versus params.host.env or args.host in doctor, lease and session-open), where the handler host parameter would be the better single source.