fix(anthropic): always sanitize empty text content blocks
Some checks failed
Unit Tests: Proxy DB Operations / proxy-db (auth-checks, tests/proxy_unit_tests/test_auth_checks.py tests/proxy_unit_tests/test_user_api_key_auth.py, 20, 8) (push) Has been cancelled
Unit Tests: Proxy DB Operations / proxy-db (key-generation, tests/proxy_unit_tests/test_key_generate_prisma.py, 30, 0) (push) Has been cancelled
Unit Tests: Proxy DB Operations / proxy-db (proxy-utils, tests/proxy_unit_tests/test_proxy_utils.py, 20, 8) (push) Has been cancelled
Unit Tests: Proxy DB Operations / proxy-db (remaining, tests/proxy_unit_tests --ignore=tests/proxy_unit_tests/test_key_generate_prisma.py --ignore=tests/proxy_unit_tests/test_auth_checks.py --ignore=tests/proxy_unit_tests/test_user_api_key_auth.py --ignore=tests/proxy_unit_tests/test_p… (push) Has been cancelled
Unit Tests: Security / security (push) Has been cancelled

Anthropic 400s on `{"role": "user", "content": ""}` with:
  "messages: text content blocks must be non-empty"

LiteLLM already had `_sanitize_empty_text_content` to rewrite empty text
to a placeholder, but it was gated behind `litellm.modify_params=True`.
With that flag off (default), empty content from upstream agent
frameworks (e.g. pydantic-ai) flowed straight through and tripped the
Anthropic validator.

Fix:
- Always run `_sanitize_empty_text_content` at the top of
  `anthropic_messages_pt`, independent of `modify_params`. There is no
  way to "pass through" an empty text block, so this is non-optional.
  The richer tool-call sanitizations (Cases A/B/D, which actually
  mutate conversation structure) remain gated on `modify_params`.
- Extend `_sanitize_empty_text_content` to also handle list-of-blocks
  content (`[{"type": "text", "text": ""}]`), not just string content.

Adds 3 regression tests covering string content, list-of-blocks
content, and the no-op case (non-empty messages with modify_params off).

Made-with: Cursor
This commit is contained in:
Krrish Dholakia 2026-04-29 19:06:10 -07:00
parent 13d9d674f9
commit 18e5aa418e
2 changed files with 147 additions and 11 deletions

View file

@ -2065,27 +2065,62 @@ def anthropic_process_openai_file_message(
)
_EMPTY_TEXT_PLACEHOLDER = (
"[System: Empty message content sanitised to satisfy protocol]"
)
def _sanitize_empty_text_content(
message: AllMessageValues,
) -> AllMessageValues:
"""
Case C: Sanitize empty text content
- Replace empty or whitespace-only text content with a placeholder message.
- Handles both string content and list-of-blocks content (rewriting only
the empty text blocks in place; non-text blocks like images are left
untouched).
Returns:
The message with sanitized content if needed, otherwise the original message
"""
if message.get("role") in ["user", "assistant"]:
content = message.get("content")
if isinstance(content, str):
if not content or not content.strip():
message = cast(AllMessageValues, dict(message)) # Make a copy
message["content"] = (
"[System: Empty message content sanitised to satisfy protocol]"
)
verbose_logger.debug(
f"_sanitize_empty_text_content: Replaced empty text content in {message.get('role')} message"
)
if message.get("role") not in ["user", "assistant"]:
return message
content = message.get("content")
if isinstance(content, str):
if not content or not content.strip():
message = cast(AllMessageValues, dict(message)) # Make a copy
message["content"] = _EMPTY_TEXT_PLACEHOLDER
verbose_logger.debug(
f"_sanitize_empty_text_content: Replaced empty text content in {message.get('role')} message"
)
return message
if isinstance(content, list):
# Walk the blocks and rewrite any empty text blocks. We rewrite (rather
# than drop) so callers don't end up with an entirely empty content
# list, which Anthropic also rejects.
new_blocks: List[Any] = []
rewrote_any = False
for block in content:
if isinstance(block, dict) and block.get("type") == "text":
text = block.get("text")
if not isinstance(text, str) or not text or not text.strip():
new_block = dict(block)
new_block["text"] = _EMPTY_TEXT_PLACEHOLDER
new_blocks.append(new_block)
rewrote_any = True
continue
new_blocks.append(block)
if rewrote_any:
message = cast(AllMessageValues, dict(message)) # Make a copy
message["content"] = new_blocks # type: ignore
verbose_logger.debug(
f"_sanitize_empty_text_content: Replaced empty text block(s) in {message.get('role')} message"
)
return message
@ -2368,6 +2403,18 @@ def anthropic_messages_pt( # noqa: PLR0915
# Sanitize messages for tool calling issues when modify_params=True
messages = sanitize_messages_for_tool_calling(messages)
# Anthropic rejects empty text content blocks with:
# "messages: text content blocks must be non-empty"
# OpenAI/other providers silently tolerate `{"role": "user", "content": ""}`,
# so callers (and upstream agent frameworks like pydantic-ai) routinely
# send empty user/assistant turns. We always rewrite these to a placeholder
# for Anthropic-shaped requests, independent of `litellm.modify_params`,
# because there is no way to "pass through" an empty text block — the
# request will always 400 otherwise. The richer tool-call sanitization
# (Cases A/B/D in `sanitize_messages_for_tool_calling`) remains gated on
# `modify_params` because it actually mutates conversation structure.
messages = [_sanitize_empty_text_content(m) for m in messages]
# add role=tool support to allow function call result/error submission
user_message_types = {"user", "tool", "function"}
# reformat messages to ensure user/assistant are alternating, if there's either 2 consecutive 'user' messages or 2 consecutive 'assistant' message, merge them.

View file

@ -339,6 +339,95 @@ class TestMessageSanitization:
assert result[0]["role"] == "user"
assert result[1]["role"] == "assistant"
def test_empty_string_content_sanitized_without_modify_params(self):
"""
Regression: An empty user message ({"role": "user", "content": ""}) must
be rewritten to a non-empty placeholder *before* it reaches Anthropic,
even when litellm.modify_params is False. Otherwise Anthropic returns:
"messages: text content blocks must be non-empty"
Reproduces a real failure from the pr-review agent (pydantic-ai).
"""
litellm.modify_params = False
messages = [
{"role": "user", "content": "First message"},
{"role": "user", "content": "please review this"},
{"role": "user", "content": ""},
]
result = anthropic_messages_pt(
messages=messages, model="claude-sonnet-4-5", llm_provider="anthropic"
)
# All three user messages get merged into one user turn for Anthropic.
assert len(result) == 1
assert result[0]["role"] == "user"
text_blocks = [
b for b in result[0]["content"] if isinstance(b, dict) and b.get("type") == "text"
]
assert len(text_blocks) == 3
# No text block may be empty — that's the contract Anthropic enforces.
for block in text_blocks:
assert block["text"].strip() != ""
assert text_blocks[2]["text"] == (
"[System: Empty message content sanitised to satisfy protocol]"
)
def test_empty_text_block_in_list_content_sanitized(self):
"""
Same regression for the list-of-blocks form:
{"role": "user", "content": [{"type": "text", "text": ""}]}
Empty text *blocks* must be rewritten too, regardless of modify_params.
"""
litellm.modify_params = False
messages = [
{
"role": "user",
"content": [
{"type": "text", "text": "real content"},
{"type": "text", "text": ""},
{"type": "text", "text": " \n "},
],
},
]
result = anthropic_messages_pt(
messages=messages, model="claude-sonnet-4-5", llm_provider="anthropic"
)
assert len(result) == 1
text_blocks = [
b for b in result[0]["content"] if isinstance(b, dict) and b.get("type") == "text"
]
assert len(text_blocks) == 3
assert text_blocks[0]["text"] == "real content"
for block in text_blocks[1:]:
assert block["text"].strip() != ""
def test_non_empty_content_unchanged_without_modify_params(self):
"""
Sanity check: when nothing is empty, the messages flow through unchanged
even with modify_params disabled.
"""
litellm.modify_params = False
messages = [
{"role": "user", "content": "Hello"},
{"role": "assistant", "content": "Hi there"},
{"role": "user", "content": "How are you?"},
]
result = anthropic_messages_pt(
messages=messages, model="claude-sonnet-4-5", llm_provider="anthropic"
)
# Two user turns + one assistant turn (alternation preserved).
assert len(result) == 3
assert result[0]["content"][0]["text"] == "Hello"
assert result[1]["content"][0]["text"] == "Hi there"
assert result[2]["content"][0]["text"] == "How are you?"
if __name__ == "__main__":
pytest.main([__file__, "-v"])