v0.3.0 Stable¶
This is the canonical public release page for the stable milestone that closes Phase 4 task-agent hardening and opens Phase 5 local scheduler foundations for the technology radar.
What This Release Proves¶
- Skylattice now has a fully inspectable collaboration tail for task-agent runs: branch push, PR preflight, PR sync, and issue-comment resume guidance all stay explicit and auditable.
- The radar now has a tracked local scheduling layer and a provider abstraction boundary without introducing a daemon or background worker.
- The public stable release signal now reflects runtime behavior again, not only packaging and authority assets.
What Works Today¶
skylattice doctor- governed
task-agentruns with deterministic edit modes, reviewed memory context, and recovery-aware GitHub collaboration sync skylattice memory ...review, search, export, and rollback flowstechnology-radarruns with GitHub discovery, provider-tagged evidence, and bounded promotion metadataskylattice radar schedule show,render, andrunfor tracked local scheduling- Windows-first CI aligned with tracked validation config
- public proof artifacts and Pages landing pages for cold visitors
Verify It¶
Why This Stable Milestone Matters¶
v0.2.2 stabilized the outward-facing authority kit. v0.3.0 moves the stable line forward by closing the operator recovery loop for GitHub collaboration and by giving radar a tracked local scheduling foundation for the next phase.
Known Limits¶
- task execution remains text-native rather than AST-aware
- meaningful task planning still depends on
OPENAI_API_KEY - radar discovery still depends on
GITHUB_TOKEN - schedule integration is Windows-first in this slice and relies on OS-level task registration
- the project is still early if you need a broad hosted automation platform