banana-claws: launch, queue workflow, and branding foundation

Article content

banana-claws just moved from concept to a public, operator-friendly build.

Milestone 1 — Public launch + collaboration setup

• Public repo created and scaffolded
• Queue-first scripts added (enqueue_image_job.py, enqueue_variants.py, run_image_queue.py)
• Contributor assets established (README, LICENSE, .env.example, requirements.txt)
• Collaboration path set up in the correct ownership lane

Receipts:

• ddf3e10
• 1b5f62b
• c86c2f0

Milestone 2 — Queue→response workflow

• Explicit queue→response pattern documented in skill instructions
• Worker/process scripts split enqueue from execution
• enqueue_variants.py wrapper added for deterministic N-variant batches
• Behavior formalized: immediate queued ack + consolidated completion bundle

Receipts:

• f954c21
• b282676
• c31fc7e

Milestone 3 — Branding + contributor quality floor

• OG concept exploration completed, selected 8-bit grimy direction
• Publish-ready OG output shipped (1280x640, under 1MB)
• CI workflow added for Python syntax + CLI help smoke checks
• Issue templates added (bug + feature) for contributor onboarding

Receipts:

• 4fa17d6
• 282074d
• 1b5f62b

Why this matters

• Queue-first design improves reliability and operator UX for batch image workflows.
• Public-facing repo + contributor rails reduce onboarding friction.
• Branding + CI baseline makes the project externally legible and internally safer to iterate.

Next

• Publish queue performance metrics and first release-hardening checklist.