docs: refresh setup and usage guide

[patlux]

Summary

Refresh the README around the actual user flow: install the provider, authenticate, select a model, and start using it.

Changes

• Move installation to the top and remove the invalid pi install pi-commandcode-provider shorthand, which pi interprets as a local path.
• Document the current pi login flow: Use a subscription → Command Code.
• Keep and clarify the unofficial integration disclaimer.
• Update the pi repository link.
• Replace repository-only model commands with commands that work after installation:
  • pi --list-models commandcode
  • omp models
• Update the Oh My Pi installation and non-interactive model examples.
• Consolidate supported authentication files and credential shapes.
• Clarify model discovery, caching, and first-offline-start behavior.
• Document the API/model-cache override environment variables.
• Remove stale promotional pricing claims.
• Clarify that a displayed zero cost for an unknown model does not mean Command Code bills it for free.
• Add update and removal commands for pi and OMP.
• Consolidate contributor and release links under development documentation.

Verification

• npm run format:check
• git diff --check
• Installed and removed the npm package through the current pi CLI.
• Verified current pi install and model-list syntax against pi 0.82.1.
• Verified OMP install, model-list, update, and uninstall syntax against the current Oh My Pi sources.
• Verified the live Command Code model endpoint and pricing documentation.
• Checked README links and workflow badges.

[github-actions]

Dependency Review

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

Scanned Files

None

[github-actions]

Memory benchmark

Compared base 46dea06 with PR head 2028322 on the same GitHub-hosted darwin runner. Lower values are better.

Metric	Base	PR	PR − Base	Change
Stable RSS	167.4 MiB ± 0.3 MiB	167.3 MiB ± 0.5 MiB	-0.7 MiB ± 0.4 MiB	-0.4%
Physical footprint	110.8 MiB ± 0.5 MiB	109.9 MiB ± 0.7 MiB	-0.9 MiB ± 0.7 MiB	-0.8%
Physical peak	126.7 MiB ± 0.5 MiB	126.8 MiB ± 0.9 MiB	-0.5 MiB ± 1.0 MiB	-0.4%

Interpretation: Differences larger than their measured MAD: Stable RSS -0.7 MiB ± 0.4 MiB, Physical footprint -0.9 MiB ± 0.7 MiB.

Extension overhead above pi baseline

Metric	Base overhead	PR overhead	Difference
Stable RSS	+5.2 MiB ± 1.3 MiB	+5.6 MiB ± 1.5 MiB	-0.7 MiB ± 0.4 MiB
Physical footprint	+3.5 MiB ± 0.3 MiB	+3.0 MiB ± 1.0 MiB	-0.9 MiB ± 0.7 MiB
Physical peak	+1.9 MiB ± 0.9 MiB	+2.7 MiB ± 1.9 MiB	-0.5 MiB ± 1.0 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.