tenant-engine/scripts/check_layer_conformance.py

#!/usr/bin/env python3
"""Check tenant-engine against the NetKingdom security layer model (§5, §11).

Read-only. Makes two mechanical checks:

1. INTENT.md frontmatter declares the layer and governs; layer.yaml is a
   derived artifact (derived: true, derived_from: INTENT.md) that must agree
   with it after ASCII case-folding. The layer is one of §3's closed four
   tokens (Taxonomy, Tooling, Engine, Staff). Neither form carries a
   standard_version (GH-DEC-2026-017 §1-§5, v0.8 amendments A9, A11, A12).
2. No catalogued Tooling client (OpenBao, key-cape) appears in src/
   unless it maps to a declared §5.1 / §5.2 / §5.3 entry.

PostgreSQL / SQLite / httpx-to-flex-auth / httpx-to-audit-core are not
Tooling contacts. They are listed in layer.yaml non_tooling_clients so
the inventory is total.

Exit 0 clean, 1 undeclared contact, 2 declaration malformed.
"""
from __future__ import annotations

import argparse
import ast
import re
import sys
from pathlib import Path

try:
    import yaml
except ImportError:  # pragma: no cover - dev extra
    print("FAIL: PyYAML is required (pip install pyyaml)", file=sys.stderr)
    raise SystemExit(2)

ROOT = Path(__file__).resolve().parents[1]
SRC = ROOT / "src" / "tenant_engine"
DECL = ROOT / "layer.yaml"
INTENT = ROOT / "INTENT.md"

# Catalogued Tooling in statute §4 today: key-cape and OpenBao.
# Import roots that would constitute a direct client of those.
TOOLING_IMPORTS = {
    "hvac": "OpenBao / Vault client",
    "bao": "OpenBao client",
    "keycloak": "key-cape / Keycloak client",
    "ldap3": "direct LDAP client (key-cape tooling)",
    "python_ldap": "direct LDAP client (key-cape tooling)",
}

TOOLING_ARGV = re.compile(r"""\[\s*(?:["']bao["']|bao_bin\b|bao_binary\b)\s*,""")
OPENBAO_ADDR = re.compile(r"\b(?:VAULT_ADDR|BAO_ADDR|X-Vault-Token)\b")

# §3 as amended (A9): closed, four tokens, compared after an ASCII case-fold.
LAYER_VOCABULARY = {"taxonomy", "tooling", "engine", "staff"}


def _fold(value: object) -> str:
    """ASCII case-fold only; two spellings of a token are one token."""
    return str(value).translate(str.maketrans("ABCDEFGHIJKLMNOPQRSTUVWXYZ",
                                              "abcdefghijklmnopqrstuvwxyz"))


def load_declaration() -> dict:
    if not DECL.exists():
        print(f"FAIL: no declaration at {DECL.relative_to(ROOT)} (§11)", file=sys.stderr)
        raise SystemExit(2)
    try:
        data = yaml.safe_load(DECL.read_text())
    except yaml.YAMLError as exc:
        print(f"FAIL: {DECL.name} is not parseable: {exc}", file=sys.stderr)
        raise SystemExit(2) from exc
    for key in ("layer", "role", "repository", "derived", "derived_from", "tooling_contacts"):
        if key not in data:
            print(f"FAIL: {DECL.name} missing required key '{key}'", file=sys.stderr)
            raise SystemExit(2)
    if data["derived"] is not True or data["derived_from"] != "INTENT.md":
        print(f"FAIL: {DECL.name} must be marked derived: true, derived_from: INTENT.md "
              "(§11, GH-DEC-2026-017 §1)", file=sys.stderr)
        raise SystemExit(2)
    if "standard_version" in data:
        print(f"FAIL: {DECL.name} carries standard_version; a layer declaration "
              "must not (GH-DEC-2026-017, A12)", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data["layer"]) not in LAYER_VOCABULARY:
        print(f"FAIL: {DECL.name} layer {data['layer']!r} is outside the closed "
              f"vocabulary {sorted(LAYER_VOCABULARY)}", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data["layer"]) != "engine":
        print(f"FAIL: declared layer is {data['layer']!r}, expected engine", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data["role"]) != "pip":
        print(f"FAIL: declared role is {data['role']!r}, expected pip", file=sys.stderr)
        raise SystemExit(2)
    if data["repository"] != "tenant-engine":
        print(f"FAIL: repository is {data['repository']!r}", file=sys.stderr)
        raise SystemExit(2)
    return data


def intent_frontmatter() -> dict:
    text = INTENT.read_text()
    if not text.startswith("---"):
        print("FAIL: INTENT.md has no YAML frontmatter", file=sys.stderr)
        raise SystemExit(2)
    block = text.split("---", 2)[1]
    data = yaml.safe_load(block) or {}
    if "layer" not in data:
        print("FAIL: INTENT.md frontmatter has no layer: key (§11)", file=sys.stderr)
        raise SystemExit(2)
    if "standard_version" in data:
        print("FAIL: INTENT.md frontmatter carries standard_version; a layer "
              "declaration must not (GH-DEC-2026-017, A12)", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data["layer"]) not in LAYER_VOCABULARY:
        print(f"FAIL: INTENT.md layer {data['layer']!r} is outside the closed "
              f"vocabulary {sorted(LAYER_VOCABULARY)}", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data["layer"]) != "engine":
        print(f"FAIL: INTENT.md layer is {data.get('layer')!r}, expected Engine", file=sys.stderr)
        raise SystemExit(2)
    if _fold(data.get("role", "")) != "pip":
        print(f"FAIL: INTENT.md role is {data.get('role')!r}, expected PIP", file=sys.stderr)
        raise SystemExit(2)
    return data


def imported_modules(path: Path) -> set[str]:
    try:
        tree = ast.parse(path.read_text())
    except SyntaxError:
        return set()
    found: set[str] = set()
    for node in ast.walk(tree):
        if isinstance(node, ast.Import):
            found.update(alias.name.split(".")[0] for alias in node.names)
        elif isinstance(node, ast.ImportFrom) and node.level == 0 and node.module:
            found.add(node.module.split(".")[0])
    return found


def scan() -> list[tuple[Path, str, str]]:
    hits: list[tuple[Path, str, str]] = []
    for path in sorted(SRC.rglob("*.py")):
        text = path.read_text()
        for module in sorted(imported_modules(path)):
            if module in TOOLING_IMPORTS:
                hits.append((path, module, TOOLING_IMPORTS[module]))
        if TOOLING_ARGV.search(text) or OPENBAO_ADDR.search(text):
            hits.append((path, "openbao-invocation", "OpenBao argv or address"))
    return hits


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--report", action="store_true")
    args = parser.parse_args()

    decl = load_declaration()
    intent = intent_frontmatter()

    # A11: a disagreement between the governing and derived forms is a finding
    # in its own right, reported rather than resolved by precedence. Case is
    # folded first (A9), so what survives is a real disagreement.
    for key in ("layer", "role"):
        if _fold(decl[key]) != _fold(intent.get(key, "")):
            print(f"FAIL: {key} disagrees after case-fold — INTENT.md (governing): "
                  f"{intent.get(key)!r}, layer.yaml (derived): {decl[key]!r}",
                  file=sys.stderr)
            return 2
    hits = scan()

    if args.report:
        print(f"tenant-engine — layer {intent['layer']}, role {intent['role']} "
              f"(INTENT.md governs; layer.yaml derived from {decl['derived_from']})")
        print(f"  tooling contacts declared: {len(decl.get('tooling_contacts') or [])}")
        print(f"  non-tooling clients: {len(decl.get('non_tooling_clients') or [])}")

    if hits:
        print("", file=sys.stderr)
        print("FAIL: undeclared Tooling-layer client (§11 undeclared violation)", file=sys.stderr)
        for path, module, what in hits:
            print(f"  {path.relative_to(ROOT)}: {module} — {what}", file=sys.stderr)
        return 1

    if decl.get("tooling_contacts"):
        print("FAIL: tooling_contacts is not empty; this engine claimed none", file=sys.stderr)
        return 1

    if not args.report:
        print(f"OK: Engine/PIP declaration matches INTENT.md; no Tooling client in "
              f"{SRC.relative_to(ROOT)}")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())