pi-commandcode-provider 0.5.1

Summary

• Add model-specific image input capabilities from the official command-code@1.15.1 catalog.
• Forward images from user messages and tool results using Command Code's current data-URL wire format.
• Advertise vision support for confirmed models including GPT-5.6 Luna and Muse Spark 1.2.
• Keep unknown and text-only models fail-closed for image input.
• Update the Command Code client version header to 1.15.1.

Contributors

• @DiyarD — reported and helped validate missing vision support in #42. Thank you!

Validation

• Full npm test suite passed.
• Typecheck and formatting passed.
• npm audit --audit-level=moderate: 0 vulnerabilities.
• Package dry-run and authenticated GPT-5.6 Luna vision E2E passed.
• CI passed for #43 and release PR #44.

Links

• Feature PR: #43
• Release PR: #44
• Issue: #42