feat: ship NDF as a Rust knowledge compiler

.github/workflows/ci.yml

@@ -6,17 +6,16 @@ on:
     branches: [main]
 
 jobs:
-  test:
+  rust:
     runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        python-version: ["3.11", "3.12"]
     steps:
       - uses: actions/checkout@v4
-      - uses: actions/setup-python@v5
+      - uses: dtolnay/rust-toolchain@1.94.0
         with:
-          python-version: ${{ matrix.python-version }}
-      - run: python -m pip install -r requirements-lock.txt
-      - run: python -m pip install -e . --no-deps
-      - run: python -m unittest discover -s tests -p 'test_*.py'
+          components: clippy,rustfmt
+      - run: cargo fmt --all -- --check
+      - run: cargo clippy --workspace --all-targets --all-features -- -D warnings
+      - run: cargo test --workspace --all-features --locked
+      - run: cargo build --workspace --release --locked
+      - run: scripts/ndf check --root tests/fixtures/markdown/project --format json
       - run: git diff --check

.gitignore

@@ -7,3 +7,4 @@ dist/
 *.sqlite
 __pycache__/
 *.py[cod]
+/target/

CHANGELOG.md

@@ -6,6 +6,7 @@ All notable changes to NDF are documented here.
 
 ### Added
 
+- Standalone Rust 1.94.0 knowledge compiler and versioned JSON CLI.
 - Canonical graph IR 0.1 with project-scoped `ndf://` identities.
 - Strict Markdown, PTO ASL, repository, test, and evidence adapters.
 - Deterministic SQLite schema v1 with semantic build hashes and provenance.
@@ -15,6 +16,8 @@ All notable changes to NDF are documented here.
 
 ### Changed
 
+- Replaced the pre-release Python prototype with a Rust-only workspace; Python
+  remains available only in Git history as migration evidence.
 - NDF format 0.2 separates `kind`, `modality`, `refinement`, `domain`, and
   `status`.
 - SQLite is defined as derived, disposable state rather than an authored
@@ -26,4 +29,5 @@ All notable changes to NDF are documented here.
   adapter.
 - PTO-SPEC ASL retains architecture authority and is normalized through the
   PTO adapter.
+- NDF has no supported Python package, import API, or PyO3 compatibility layer.
 - Format, IR, JSON output, CLI, and plugin surfaces remain experimental.