verify(P10): contract-ingestor-defense-in-depth — 4-layer verify PASS + ship

VERIFY: structural — fail-closed + env discovery; behavioral — 49 tests + CI PASS; security — defense-in-depth on IAM identity.

---ci---
project: acdl
phase: 10
milestone: v1.16
status: complete
phase_role: execution
requirements:
  covered: [REQ-174]
  partial: []
---/ci---

core/lambda/contract_ingestor.py

@@ -34,6 +34,21 @@ _dynamodb = None
 _secrets_client = None
 
 
+def _discover_environments():
+    """P10 (REQ-174): derive the valid environment names from
+    core/environments/*.json (the directory is the single source of truth,
+    not a hardcoded set). Falls back to {'dev','qa','prod','dr'} if the
+    directory is not readable (e.g. packaged Lambda without the dir).
+    """
+    env_dir = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(
+        os.path.abspath(__file__)))), "core", "environments")
+    try:
+        names = {f[:-5] for f in os.listdir(env_dir) if f.endswith(".json")}
+        return names or {"dev", "qa", "prod", "dr"}
+    except OSError:
+        return {"dev", "qa", "prod", "dr"}
+
+
 def _get_dynamodb():
     global _dynamodb
     if _dynamodb is None:
@@ -236,20 +251,33 @@ def _validate_caller_identity(event, payload):
     in the payload matches the principal's ARN-derived source identity, preventing
     one consumer from impersonating another.
 
-    If the identity is not available (e.g. local testing or non-IAM auth), the
-    check is skipped (the ABAC policy at the IAM layer enforces the scope).
+    P10 (REQ-174): if the IAM identity is absent (no callerArn), the function
+    FAILS CLOSED (raises ValueError) rather than silently passing. The ABAC
+    policy at the IAM layer is the primary enforcement; this is defense-in-
+    depth so a misconfigured Function URL (no IAM auth) does not allow
+    unauthenticated contract submission. Local testing must set a test ARN
+    via the event requestContext or the LOCAL_LAMBDA_STUB env bypass.
 
     v1.14 (REQ-144): also validates contractId format, environment enum, and
-    error length. The ABAC reliance is documented here: the Function URL IAM
-    identity does not expose principal tags in the event, so full enforcement
-    of consumerRepo ownership is at the IAM layer (ABAC via
-     aws:PrincipalTag/nova:owner). This function validates format only, not
-    ownership.
+    error length. P10 (REQ-174): the environment enum is derived from the
+    core/environments/ directory (not hardcoded), so a new env JSON is the
+    single source of truth. The ABAC reliance is documented here: the
+    Function URL IAM identity does not expose principal tags in the event,
+    so full enforcement of consumerRepo ownership is at the IAM layer (ABAC
+    via aws:PrincipalTag/nova:owner). This function validates format only,
+    not ownership.
     """
     identity = event.get("requestContext", {}).get("identity", {})
     caller_arn = identity.get("userArn", "")
     if not caller_arn:
-        pass  # no identity available — rely on IAM ABAC enforcement
+        # P10 (REQ-174): fail closed. A local-test bypass is allowed via
+        # the NOVA_LAMBDA_LOCAL_BYPASS env var (set by the LocalLambdaStub).
+        import os as _os
+        if not _os.environ.get("NOVA_LAMBDA_LOCAL_BYPASS"):
+            raise ValueError(
+                "missing IAM caller identity (requestContext.identity.userArn) — "
+                "the Function URL must use IAM auth; refusing unauthenticated submission"
+            )
     payload_repo = payload.get("consumerRepo", "")
     if payload_repo:
         # consumerRepo must be org/repo format, <=128 chars
@@ -263,12 +291,13 @@ def _validate_caller_identity(event, payload):
         if not re.match(r'^[a-zA-Z0-9][a-zA-Z0-9_-]{0,63}$', contract_id):
             raise ValueError(f"invalid contractId format: {contract_id!r} (alphanumeric, hyphen, underscore; max 64 chars)")
 
-    # v1.14 (REQ-144): environment enum validation
+    # P10 (REQ-174): environment enum derived from core/environments/ (not
+    # hardcoded) — the directory is the single source of truth.
     environment = payload.get("environment", "")
     if environment:
-        valid_envs = {"dev", "qa", "prod", "dr"}
+        valid_envs = _discover_environments()
         if environment not in valid_envs:
-            raise ValueError(f"invalid environment: {environment!r} (must be one of {valid_envs})")
+            raise ValueError(f"invalid environment: {environment!r} (must be one of {sorted(valid_envs)})")
 
     # v1.14 (REQ-144): error length cap (for report_error action)
     error_msg = payload.get("error", "")
@@ -357,6 +386,9 @@ def lambda_handler(event, context):
             }
         return {"statusCode": 200, "body": json.dumps(result)}
     except ValueError as e:
+        # P10 (REQ-174): identity failures are 401, field validation is 400.
+        if "missing IAM caller identity" in str(e):
+            return {"statusCode": 401, "body": json.dumps({"error": str(e)})}
         return {"statusCode": 400, "body": json.dumps({"error": str(e)})}
     except Exception as e:  # pragma: no cover - defensive top-level guard
         return {"statusCode": 500, "body": json.dumps({"error": str(e)})}