Skip to content

Latest commit

 

History

History
12 lines (9 loc) · 1.19 KB

File metadata and controls

12 lines (9 loc) · 1.19 KB

Grinta Showcase

These case studies make a narrow claim: they document what Grinta did in recorded runs and distinguish runtime completion from broader specification compliance. Exact artifacts are linked wherever they are public; unavailable or private evidence is labeled as such.

Case study Result Evidence
4h 33m autonomous session 16,393 events, 373 tool outcomes, one initial user turn, final state FINISHED Sanitized report and committed audit summary
Failure recovery Failure output inspected, targeted repair applied, validation rerun without a follow-up prompt Demo recording and disclosure notes
Issue-tracker build Flask + React issue tracker produced and validated with 19 passing tests Run summary; public transcript pending sanitization
Raft key-value store Race-condition failure diagnosed; 39/39 tests passed consistently Full demo and animated preview

For the evidence standard used by the engineering narrative, see docs/journey/EVIDENCE.md.