We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
macOS claude/gemini NI PASS, macOS claude/kimi NI PASS
Windows codex/kimi NI FAIL
Ubuntu pi/mini NI PASS, Windows codex/sonnet NI FAIL
macOS claude/gpt-5.4-mini NI PASS — completes gpt-5.4-mini NI row for claude+codex
Add pi/gpt-5.4-mini NI: macOS PASS, Windows FAIL
Ubuntu claude/mini NI PASS (was git flake), Windows codex/gemini NI PASS
Update: Windows sonnet NI PASS (was FAIL), macOS gemini NI PASS, claude/mini FAIL evidence
Add pi/gpt-5.5 BI PASS + NI FAIL evidence
Add 19 FAIL evidences to QA matrix — vanilla NI/BI failures across platforms
Add Windows codex/gpt-5.4-mini BI PASS
Add Windows gpt-5.4-mini NI PASS for claude-code + codex
Add 4 more PASS: macOS sonnet NI, Windows gpt-5.5 BI, Ubuntu gemini BI
Add 6 new PASS evidences: gpt-5.4-mini NI+BI, gemini NI, sonnet BI, pi/kimi BI
Fill historical evidence: claude/sonnet NI, claude/kimi NI, codex/gemini NI, bp/predefined, bp/create passes from May 21-23
Add gpt-5.4-mini model to all QA matrix sections
Add claude-sonnet-4-6 and gemini-3.5-flash model sections to all BP test types
Complete QA matrix: all agents x all models x all modes x all BP types
Add Kimi-K2.6 rows to all BP test sections
Add BP interactive + bridged-hooks modes to QA matrix
Expand QA evidence with all agents (pi, hermes, cursor, copilot, opencode, gemini-cli) and models (sonnet-4-6, gemini, kimi)
Update QA evidence with live-stack pass results (2026-05-23)
Created QA Evidence (markdown)
docs: daily update 2026-05-21
Initial Home page