Skip to content

Document conversational photo-to-robot skill workflow and runtime evidence - #47

Open
Tarzelf wants to merge 2 commits into
mainfrom
codex/chat-to-robot-skills-20261027
Open

Tarzelf wants to merge 2 commits into
mainfrom
codex/chat-to-robot-skills-20261027

Conversation

@Tarzelf

@Tarzelf Tarzelf commented Oct 28, 2026 •

Copy link
Copy Markdown
Contributor

Connects the photo-and-outcome concept to a conversational workflow for compatible robot skills, verified installation, real simulation and action/activation inspection. The report distinguishes the implemented local tool boundary from general household planning and hardware qualification.

Publishes the original SSDlite photo-inference receipt, actual repeatable LIBERO task-5 reset, pinned upstream revisions and a documentation-only skill manifest. OpenVLA weights remain incomplete: no learned rollout, task completion or steering benefit is claimed.

The matching Hugging Face documentation is already published. The publication receipts verify uploaded bytes anonymously and preserve the existing Space evidence. The final revision includes the verified desktop chat screenshot.

Validation: npm run check passes; all 75 research tests pass using node --test --test-concurrency=1 scripts/*.test.mjs. Source and generated readers are included. No private website source, credentials, weights or participant photos are published.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant