Changelog
v0.3.0 — agentura generate
March 2026
- New: agentura generate command
- Generate 15 eval test cases from a plain-English description
- Optional live agent probing for better generation quality
- Auto-generates quality rubric for llm_judge
- Updates agentura.yaml with all 3 eval strategies
- Powered by Groq LLaMA 3.3 70B (free API)
v0.2.0 — CLI + Docs
March 2026
- New: agentura CLI (init, run, login)
- New: Browser-based authentication flow
- New: API key management in dashboard
- New: /docs pages with quickstart and reference
v0.1.0 — Initial Release
March 2026
- GitHub App integration
- Three eval strategies: golden_dataset, llm_judge, performance
- PR comments with results table
- GitHub Check Run integration
- Baseline comparison against main branch
- Web dashboard