fix(chat): detect CONTINUATION conversation history headers

[syf2211]

Summary

Fix duplicate conversation history reconstruction when the server injects a (CONTINUATION) header during chat continuations.

Motivation

server.py embeds history using build_conversation_history(), which emits:

=== CONVERSATION HISTORY (CONTINUATION) ===

SimpleTool only checked for the exact legacy sentinel === CONVERSATION HISTORY ===, so continuation prompts with the server header were treated as missing history. The tool then rebuilt the thread again, duplicating context and inflating prompt size.

Fixes #459

Changes

• Add has_embedded_conversation_history() in utils/conversation_memory.py
• Use the helper in tools/simple/base.py when deciding whether to reuse pre-embedded history
• Add regression tests for continuation and legacy header formats

Tests

pytest tests/test_conversation_continuation_integration.py -q

Result: 4 passed

Notes

• Detection requires both the start prefix (=== CONVERSATION HISTORY) and end marker (=== END CONVERSATION HISTORY ===) to avoid false positives.
• No runtime behavior change for in-process continuation paths that legitimately need reconstruction.

[gemini-code-assist]

Code Review

This pull request introduces the has_embedded_conversation_history helper function to check if a prompt contains server-built conversation history, replacing a hardcoded string check in tools/simple/base.py and adding corresponding integration tests. The review feedback suggests making this helper function more robust by handling None or non-string inputs gracefully and adding tests for these edge cases.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

The function has_embedded_conversation_history assumes that prompt is always a valid string. However, if prompt is None or of another non-string type (which can happen if a tool's request prompt is optional or unset), performing 'in prompt' operations will raise a TypeError at runtime.

To enforce defensive programming and prevent potential crashes, we should add a type and null check at the beginning of the function.

Suggested change

	def has_embedded_conversation_history(prompt: str) -> bool:
	"""Return True when a prompt already contains server-built conversation history."""
	return (
	"=== CONVERSATION HISTORY" in prompt
	and "=== END CONVERSATION HISTORY ===" in prompt
	)
	def has_embedded_conversation_history(prompt: Optional[str]) -> bool:
	"""Return True when a prompt already contains server-built conversation history."""
	if not prompt or not isinstance(prompt, str):
	return False
	return (
	"=== CONVERSATION HISTORY" in prompt
	and "=== END CONVERSATION HISTORY ===" in prompt
	)

[gemini-code-assist]

It is a good practice to add test cases for edge cases such as None or non-string inputs to verify that has_embedded_conversation_history handles them gracefully without raising exceptions.

Suggested change

	def test_has_embedded_conversation_history_rejects_plain_prompt():
	assert has_embedded_conversation_history("continue the discussion") is False
	def test_has_embedded_conversation_history_rejects_plain_prompt():
	assert has_embedded_conversation_history("continue the discussion") is False
	assert has_embedded_conversation_history(None) is False