Skip to content

lablink-cli v0.3.0

Choose a tag to compare

View all tags
@7174Andy 7174Andy released this 08 Sep 17:48
· 6 commits to main since this release
lablink-cli_v0.3.0
872a161
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Changes

Features

  • lablink deploy prompts (hidden input) for a Tailscale authkey or Cloudflare Tunnel token when the flag is omitted, instead of refusing and forcing the credential onto the command line (#485)
  • lablink doctor warns when an ssl.provider: none (HTTP) deployment silently disables H.264 — browsers expose WebCodecs only on secure origins, so sessions fall back to JPEG/WebP stills and feel laggy (#491)

Bug Fixes

  • lablink doctor's AMI check verifies the image exists in the target region instead of trusting AMI_MAP, whose false green ended in InvalidAMIID.NotFound at deploy time (#484)
  • An empty machine.ami_id resolves the per-region Deep Learning Base AMI fallback (matching the deployment behavior from #489) instead of failing doctor (#490)

Changed

  • Pinned lablink-template bumped to v0.3.1: app.region now actually drives the deployment, and the allocator AMI resolves per region via SSM (#498)

Installation

uv tool install lablink-cli   # or: pip install lablink-cli==0.3.0

Upgrade notes

  • Requires lablink-allocator-service>=0.4.0 (released today): the CLI re-exports MachineConfig, whose ami_id default is now empty = per-region AMI fallback. The pin enforces the pairing.
  • Deploy credentials now need ssm:GetParameter on /aws/service/canonical/* (template v0.3.1) — see docs/aws-setup.md.
  • If your config names a region other than us-west-2, destroy the old deployment first: existing us-west-2 resources are not moved and keep billing.
Assets 2
Loading