Show HN: Primer – build software with AI agents one milestone at a time (github.com)

[dead]

Have you seen openspec, or any of the others in this space?

[flagged]

Planning and specs are what makes AI capable and reliable

How do you think Primer will evolve as agents get better at long-horizon tasks? The milestone and verification loop make sense now as agents fail unpredictably on complex tasks. Do tasks just get bigger and milestones scale with them?