Skip to content

v3.13.0: QE-Court — adversarial review as a verdict

Choose a tag to compare

View all tags
@proffesor-for-testing proffesor-for-testing released this 18 Jul 18:20
v3.13.0
001c042
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

What's New

QE-Court — an adversarial review court where a change is prosecuted by independent AI reviewers from different vendors (Codex/GPT, Cognitum, Claude), each with its own probe set, and a SHIP verdict must survive an escalating deeper review before it stands (SHIP / REMAND / BLOCK). It debuted by reviewing this release and blocked it, catching three data-loss/crash bugs a green test suite had missed — all fixed.

Also in this release:

  • Codex (GPT) provider — run QE on your ChatGPT subscription.
  • Statistical significance gate for the self-learning flywheel — no more promoting within-noise gains.
  • Automatic memory.db recovery from verified backups on genuine corruption (never touches a healthy DB).
  • Per-agent cost attribution in the spend ledger.
  • 4 HIGH-severity CVEs cleared from consumer installs (#565).

Getting Started

```bash
npx agentic-qe init --auto
```

Try the court on your own changes: `/qe-court` (or `aqe eval run --skill qe-court`).

See CHANGELOG and release notes for full details.

Assets 2
Loading