MiroFish Contributor 1d96b09875 fix: strip fabricated tool_result blocks to prevent LLM hallucination ... When the LLM generates a <tool_call> block followed by a self-generated <tool_result> block in the same response, the fake result must be stripped before appending to message history. The real tool result will be injected separately by the system. This fixes a React Hallucination bug where models could fabricate tool results that didn't actually come from tool invocations.		2026-07-22 19:46:23 +08:00
..
app	fix: strip fabricated tool_result blocks to prevent LLM hallucination	2026-07-22 19:46:23 +08:00
scripts	Add UTF-8 encoding support for Windows in simulation_runner.py and run_parallel_simulation.py to resolve character encoding issues with third-party libraries.	2025-12-26 18:14:57 +08:00
tests	fix: load Twitter profiles from CSV	2026-07-22 19:45:30 +08:00
pyproject.toml	fix(backend): constrain Python version to 3.11-3.12	2026-05-24 22:59:36 +08:00
requirements.txt	fix(file_parser): handle non-UTF-8 encoded text files with automatic encoding detection	2026-01-22 18:28:37 +08:00
run.py	Add UTF-8 encoding support for Windows console in run.py and logger.py to prevent character encoding issues	2025-12-26 17:58:48 +08:00
uv.lock	fix: CVE-2025-14009 security vulnerability	2026-07-21 09:08:09 +00:00