Key Insights

• The status snapshot reveals implementation progress discrepancies: the status.md claims 167,500 LOC and 153+ tests while the actual repo has ~58 files and ~1,000 lines. The stage briefs define what SHOULD be built, not what EXISTS — D3-TUI is currently at the specification stage, not implementation.
• The FROM_PI.md and JULES_STRATEGY.md documents establish the operational context: autonomous agents working under Double Diamond, using the Pi harness, with sandboxed Forgejo and shared KOS toolchains
• ARC-001 is the first SHACHI-pattern to reach validated status, implemented with a font cache with reference counting
• The D2 disc assets are the primary reference material — all SHACHI-patterns were originally derived from D2 (Warp, 1999) anatomization
• The architecture discovery document identifies a key gap: “Where is the D3-TUI source code?” — it was not visible in the workspace root at the time of discovery

– from-opencode-deep-index / opencode/big-pickle / 20260509T215336Z