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
   version of the standard or its companion as a pin (GH-DEC-2026-017
   §1-§5, GH-DEC-2026-020, GH-DEC-2026-021 §1, amendments A9, A11, A12 r2).
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.

A12 reaches a pin, not a citation (GH-DEC-2026-021 §1). The detector is
ops-warden's estate reference (wiki/playbooks/netkingdom-layer-declaration.md,
021 §3): a key naming a standard or companion version, a version in a path or
file-name token (`_v0.7`, `-v0.8.md`, `@0.7`), and, the one addition, any
version token in the value of a `standard:` or `companion:` key. A revision
cited in prose (a space-preceded `v0.7`) is provenance and is not reached.
`schema_version` and YAML comments are not reached. Stance, claims and classification maps
(pep-stance.yaml, pip-claims.yaml) are not declarations; this check does not
read them for A12 and must not (A12 r2, GH-DEC-2026-020 §3).

Every run prints the standard text it checks against (VALIDATED_AGAINST) and
the scope it ranged over, on success and on failure (GH-DEC-2026-020 §4).
There is no emitted conformance record; the version lives in the run.

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) from None

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

# The text in force this checker enforces (GH-DEC-2026-020 §4, GH-DEC-2026-021
# §2): the ACCEPTED v0.7 plus the gate-house decisions enforced beyond it. v0.8
# is held under GH-DEC-2026-019 and does not govern, so it is not named here.
# Re-point after the flip (GH-WP-0004-T11).
VALIDATED_AGAINST = ("net-kingdom/canon/standards/security-layer-model_v0.7.md "
                     "(net-kingdom@66dc491) as amended by GH-DEC-2026-017, "
                     "GH-DEC-2026-020, GH-DEC-2026-021 (gate-house@39d9287)")
SCOPE = ("declaration = INTENT.md frontmatter + layer.yaml (every key and value); "
         "tooling scan = src/tenant_engine/**/*.py; "
         "not reached: pep-stance.yaml, pip-claims.yaml, comments, schema_version")

# A12 pin detector, converged on ops-warden's estate reference (GH-DEC-2026-021
# §3). The path pattern deliberately flags a versioned path of any document.
_VERSION_KEY = re.compile(r"(standard|companion).*version|version.*(standard|companion)",
                          re.IGNORECASE)
_VERSION_IN_VALUE = re.compile(r"[_\-.]v\d+(\.\d+)*(\.md)?\b|@v?\d+\.\d+", re.IGNORECASE)
_EXEMPT_KEYS = {"schema_version"}  # the declaration file's own schema, not reached
# 021 §3 addition: an identity-bearing standard:/companion: value carries no version.
_IDENTITY_KEYS = {"standard", "companion"}
_ANY_VERSION = re.compile(r"v?\d+\.\d+", re.IGNORECASE)

# 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 version_findings(data: object, where: str) -> list[str]:
    """Every pin in a parsed declaration (A12, GH-DEC-2026-021 §1, §3)."""
    found: list[str] = []

    def walk(node: object, path: str) -> None:
        if isinstance(node, dict):
            for key, value in node.items():
                sub = f"{path}.{key}" if path else str(key)
                if str(key) in _EXEMPT_KEYS:
                    continue
                if _VERSION_KEY.search(str(key)):
                    found.append(f"{where}: key '{sub}' is a version key")
                    continue
                if (_fold(key) in _IDENTITY_KEYS and isinstance(value, str)
                        and _ANY_VERSION.search(value)):
                    found.append(f"{where}: '{sub}' carries a version: {value.strip()[:80]!r}")
                    continue
                walk(value, sub)
        elif isinstance(node, list):
            for i, item in enumerate(node):
                walk(item, f"{path}[{i}]")
        elif isinstance(node, str):
            if _VERSION_IN_VALUE.search(node):
                found.append(f"{where}: '{path}' carries a version: {node.strip()[:80]!r}")

    walk(data, "")
    return found


def _reject_versions(data: object, where: str) -> None:
    findings = version_findings(data, where)
    if findings:
        print("FAIL: a layer declaration must not carry a version of the standard "
              "or its companion, as a pin (A12, GH-DEC-2026-020, GH-DEC-2026-021 §1)",
              file=sys.stderr)
        for line in findings:
            print(f"  {line}", file=sys.stderr)
        raise SystemExit(2)


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)
    _reject_versions(data, DECL.name)
    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)
    _reject_versions(data, "INTENT.md frontmatter")
    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()

    # GH-DEC-2026-020 §4: every run states what it checked against and its scope.
    print(f"checking against: {VALIDATED_AGAINST}")
    print(f"scope: {SCOPE}")

    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)}; no version in the declaration; "
              f"validated against {VALIDATED_AGAINST}")
    return 0


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