@@ -29,7 +29,8 @@ <h1>QRAE-RD V0</h1>
2929 < p >
3030 QRAE-RD is a human-gated, agent-orchestrated quant research operating system.
3131 It uses deterministic code for data validation, backtesting, walk-forward analysis,
32- risk scoring, Research Director packets, allocator-style reporting, and requirement-level traceability.
32+ risk scoring, Research Director packets, allocator-style reporting, quality gates,
33+ and requirement-level traceability.
3334 </ p >
3435 < div class ="grid ">
3536 < section class ="card ">
@@ -52,6 +53,11 @@ <h2>Director packet</h2>
5253 < p > Fetch the survivor packet for ChatGPT Plus, Gemini CLI, or future API Research Director review.</ p >
5354 < a href ="/api/packets/example "> Open /api/packets/example</ a >
5455 </ section >
56+ < section class ="card ">
57+ < h2 > Quality gates</ h2 >
58+ < p > Review regression test cases, fatal failures, scoring weights, and output acceptance thresholds.</ p >
59+ < a href ="/api/quality/gates "> Open /api/quality/gates</ a >
60+ </ section >
5561 < section class ="card ">
5662 < h2 > Traceability audit</ h2 >
5763 < p > Map every design requirement from the project chat to implementation, docs, live endpoints, or V1 deferral.</ p >
@@ -62,7 +68,7 @@ <h2>Traceability audit</h2>
6268 V0 deliberately disables live trading and automated paid LLM execution. Metrics must come from artifacts,
6369 robustness must come from WFO, and capital approval requires a human gate.
6470 </ p >
65- < p > < code > scripts verify → candidates survive → Research Director reviews → human approves</ code > </ p >
71+ < p > < code > scripts verify → candidates survive → quality gates check → Research Director reviews → human approves</ code > </ p >
6672 </ main >
6773</ body >
6874</ html >
0 commit comments