fix(hide-secrets): scan the first token of shell-style assignments regardless of what follows

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

enterprise/litellm_enterprise/enterprise_callbacks/secret_detection.py

@@ -451,9 +451,9 @@ _CONFIG_SECTION: Final = "litellm-prompt"
 
 _ASSIGNMENT_LINE: Final = re.compile(r"[^\s\[#;:=][^:=]*[:=]")
 
-_SHELL_OPERATORS: Final = ";&|"
+_SHELL_ASSIGNMENT: Final = re.compile(r"(?P<key>[^\s\[#;:=](?:[^:=]*[^\s:=])?)=(?P<value>\S+)")
 
-_SHELL_TRAILER: Final = re.compile(rf"\\|[{_SHELL_OPERATORS}]+|#.*|\S+=\S*")
+_SHELL_OPERATORS: Final = ";&|"
 
 _SCAN_SUFFIX: Final = ".py"
 
@@ -491,6 +491,9 @@ def _classify_line(state: tuple[bool, str | None], numbered: tuple[int, str]) ->
     stripped: Final = line.strip()
     if not stripped or stripped[0] in "#;":
         return open_option, None
+    shell_assignment: Final = _SHELL_ASSIGNMENT.match(stripped)
+    if shell_assignment is not None:
+        return True, f"{shell_assignment['key']}_{number}={shell_assignment['value']}"
     assignment: Final = _ASSIGNMENT_LINE.match(stripped)
     if assignment is not None:
         return True, f"{assignment.group()[:-1].strip()}_{number}{stripped[assignment.end() - 1 :]}"
@@ -505,14 +508,10 @@ def _parseable_lines(text: str) -> Iterator[str]:
 
 
 def _lone_value(line: str) -> str | None:
-    tokens: Final = line.split(maxsplit=2)
-    if not tokens or '"' in tokens[0]:
+    tokens: Final = line.split()
+    if not tokens or '"' in tokens[0] or (len(tokens) > 1 and not tokens[1].startswith("#")):
         return None
-    value: Final = tokens[0].rstrip(_SHELL_OPERATORS)
-    trailer: Final = tokens[1] if len(tokens) > 1 else ""
-    if value != tokens[0] or not trailer or _SHELL_TRAILER.fullmatch(trailer) is not None:
-        return value
-    return None
+    return tokens[0].rstrip(_SHELL_OPERATORS)
 
 
 def _quoted_assignments(text: str) -> tuple[str, ...]:

tests/test_litellm/enterprise/enterprise_callbacks/test_secret_detection.py

@@ -120,6 +120,9 @@ def test_scan_message_preserves_quoted_benign_identifiers():
         ("export DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt DB_HOST=db.internal", "Zx4Kp9Lm2Qr7Ns3Vt"),
         ("DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt; systemctl restart app", "Zx4Kp9Lm2Qr7Ns3Vt"),
         ("DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt | tee creds.txt", "Zx4Kp9Lm2Qr7Ns3Vt"),
+        ("DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt > setup.log", "Zx4Kp9Lm2Qr7Ns3Vt"),
+        ("docker run -e DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt --name app postgres", "Zx4Kp9Lm2Qr7Ns3Vt"),
+        ("password=correcthorsebattery please", "correcthorsebattery"),
     ],
     ids=[
         "env-password",
@@ -153,6 +156,9 @@ def test_scan_message_preserves_quoted_benign_identifiers():
         "second-assignment-after-the-value",
         "semicolon-after-the-value",
         "pipe-after-the-value",
+        "redirect-after-the-value",
+        "docker-flag-after-the-value",
+        "prose-after-a-shell-assignment",
     ],
 )
 def test_scan_message_redacts_credentials_assigned_to_credential_keys(content, secret):
@@ -171,6 +177,16 @@ def test_scan_message_redacts_only_the_first_token_of_a_shell_assignment():
     )
 
 
+@pytest.mark.parametrize("operator", [";", "&&", "|"])
+def test_scan_message_keeps_a_shell_operator_glued_to_the_value(operator):
+    guardrail = _guardrail()
+
+    assert (
+        guardrail.redact_text(f"DB_PASSWORD=Zx4Kp9Lm2Qr7Ns3Vt{operator} systemctl restart app")
+        == f"DB_PASSWORD=[REDACTED]{operator} systemctl restart app"
+    )
+
+
 def test_scan_message_closes_a_yaml_block_at_the_next_unindented_line():
     guardrail = _guardrail()
     content = "api_key: >\n  aB3dE6gH9jK2mN5p\nSteps\n  Rotate-Before-Friday please"
@@ -244,6 +260,7 @@ def test_scan_message_redacts_every_credential_on_one_line():
         "password_hint: your usual one followed by Ticket-LIT7049-Suffix",
         "Translate this recipe note into French:\nsecret_sauce: Worcestershire sauce",
         "api_key = Massachusetts (the state, not a key)",
+        "secret_sauce:Worcestershire sauce",
     ],
     ids=[
         "prose-password",
@@ -300,6 +317,7 @@ def test_scan_message_redacts_every_credential_on_one_line():
         "sentence-holding-a-later-mixed-case-token",
         "capitalized-word-starting-a-phrase",
         "capitalized-word-before-a-parenthetical",
+        "yaml-scalar-without-a-space-after-the-colon",
     ],
 )
 def test_scan_message_keeps_benign_values(content):