docs: add unified levels documentation

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: hyperpolymath

UNIFIED-LEVELS.md

@@ -0,0 +1,57 @@
+// SPDX-License-Identifier: PMPL-2.0-or-later
+= Unified Level System — TypeLL / VQL-UT / VeriSimDB / TypedQLiser / PanLL
+:toc:
+
+== Level Definitions
+
+[cols="1,3,3"]
+|===
+| Level | Name | Criteria
+
+| L0 | Scaffold | Repo exists, no functional code
+| L1 | Spec | Specifications/types defined, no runtime
+| L2 | Partial | Some code works standalone, not connected
+| L3 | Core | Main engine works, tests pass, standalone
+| L4 | Integrated | Connected to adjacent systems, data flows
+| L5 | End-to-end | Full pipeline works across all systems
+| L6 | Tested | Point-to-point, end-to-end, aspect tests + benchmarks
+| L7 | Production | Published, documented, CI green, ready for users
+|===
+
+== Current State (2026-03-21)
+
+[cols="2,1,3"]
+|===
+| Component | Level | Notes
+
+| TypeLL core engine | L1 | Specs exist, no runtime type checker
+| TypedQLiser L5-10 | L1 | Levels 1-4 work, 5-10 need TypeLL
+| PanLL VQL panel | L0 | Does not exist
+| VQL-UT → VeriSimDB bridge | L2 | Parser works, no DB execution
+| VeriSimDB octad storage | L4 | Phase 4 validated, production tested
+| PanLL TypeLL panel | L2 | ReScript UI exists, not verified running
+| TypeLL-VQL bridge | L2 | Crate exists, depends on typell-core
+|===
+
+== Level-Up Plan
+
+Round 1: Bring trailing (L0/L1) → L3::
+  1. TypeLL core engine: implement type checker runtime
+  2. PanLL VQL panel: create the panel
+  3. TypedQLiser L5-10: implement with TypeLL backing
+
+Round 2: Bring all → L4 (integrated)::
+  4. Wire VQL-UT → VeriSimDB runtime bridge
+  5. Wire TypeLL → TypedQLiser for levels 5-10
+  6. Wire PanLL TypeLL panel → running TypeLL server
+  7. Wire PanLL VQL panel → VQL-UT
+
+Round 3: Bring all → L5 (end-to-end)::
+  8. PanLL → TypeLL → VQL-UT → VeriSimDB pipeline test
+  9. TypedQLiser → TypeLL → VQL-UT type checking pipeline
+
+Round 4: Bring all → L6 (tested + benchmarked)::
+  10. Point-to-point tests for every connection
+  11. End-to-end tests for full pipelines
+  12. Aspect tests (error handling, edge cases, concurrency)
+  13. Benchmarks (latency, throughput, memory)