feat(tests): add isolated pi launchers

[patlux]

Summary

• add a disposable logged-out launcher for testing the current checkout and /login
• add an authenticated launcher that uses existing Command Code auth files and limits model selection to Command Code
• resolve stored Command Code credentials before provider registration for pi 0.84 compatibility
• document both development workflows and cover them with launcher tests

Verification

• npm test
• npm run format:check
• git diff --check
• env -u COMMANDCODE_API_KEY pi --no-extensions -e ./index.ts --list-models commandcode
• interactive /login and authenticated Command Code request confirmed by user

[github-actions]

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Scanned Files

None

[github-actions]

Memory benchmark

Compared base 59d133e with PR head 62e51ee on the same GitHub-hosted darwin runner. Lower values are better.

Metric	Base	PR	PR − Base	Change
Stable RSS	167.8 MiB ± 0.6 MiB	167.6 MiB ± 1.4 MiB	+0.0 MiB ± 1.8 MiB	+0.0%
Physical footprint	110.6 MiB ± 0.1 MiB	110.7 MiB ± 0.5 MiB	-0.1 MiB ± 0.4 MiB	-0.1%
Physical peak	127.1 MiB ± 0.2 MiB	127.5 MiB ± 1.4 MiB	+0.4 MiB ± 1.8 MiB	+0.3%

Interpretation: No metric shows a clear Base-to-PR difference beyond its measured run-to-run variation.

Extension overhead above pi baseline

Metric	Base overhead	PR overhead	Difference
Stable RSS	+6.9 MiB ± 0.2 MiB	+7.4 MiB ± 0.8 MiB	+0.0 MiB ± 1.8 MiB
Physical footprint	+4.9 MiB ± 0.1 MiB	+5.1 MiB ± 0.2 MiB	-0.1 MiB ± 0.4 MiB
Physical peak	+4.2 MiB ± 0.5 MiB	+5.5 MiB ± 0.8 MiB	+0.4 MiB ± 1.8 MiB

Values are medians of 6 alternating, paired runs. The value after ± is the median absolute deviation (MAD). Each process was sampled 12 times after a 2500 ms warm-up.

Environment: pi 0.82.1, Bun 1.3.11, darwin arm64. RSS comes from ps; physical footprint and peak come from macOS footprint.

This is a comparative signal, not a pass/fail threshold. GitHub-hosted runner noise can affect absolute values.