对于每个智能体 a，算法统计它及其谱系的成功与失败次数：

然后基于这些统计，使用 Thompson Sampling 抽样策略选出下一个扩张对象：

评估与扩张的过程被设计为异步并行：每个 CPU 核心独立运行，当某个任务结束时立即补位新任务，从而持续利用全部算力。

最终，算法在预算耗尽后通过正则化不完全 Beta 函数  选出“最可信”的智能体：

+def attempt_error_resolution(git_dir, test_output, test_error,
,→ language):
+ """
+ Attempt to automatically diagnose and resolve errors.
+ Returns a tuple of (resolved, message) where resolved indicates if
,→ errors were fixed.
+ """
+ safe_log("Attempting automated error diagnosis and resolution...")
+
+ # Diagnose errors using our enhanced bash tool function
+ diagnosis = diagnose_errors(test_output, test_error, "")
+
+ if not diagnosis["has_errors"]:
+ return False, "No errors detected to resolve."
+
+ resolution_messages = []
+
+ # Try to apply automated fixes for each diagnosed error
+ for error in diagnosis["errors"]:
+ safe_log(f"Processing error: {error[’type’]} -
,→ {error[’description’]}")
+
+ # Simple resolution strategies based on error type
+ if error["type"] == "python_module_not_found":
+ # For Python module not found errors, we might install the
,→ module
+ match = re.search(r"No module named ’([^’]+)’",
,→ error["description"])
+ if match:
+ module = match.group(1)
+ resolution_messages.append(f"Would attempt to install
,→ Python module: {module}")
+ # In practice, we would run: pip install {module}
+ # But we’ll skip actual installation to avoid side
,→ effects
+
+ elif error["type"] == "python_syntax_error" and "file"in error:
+ # For syntax errors, we could potentially apply fixes
+ file_path = os.path.join(git_dir, error["file"])
+ if os.path.exists(file_path):
+ resolution_messages.append(f"Would attempt to fix
,→ syntax error in {file_path} at line {error.get(’line’,
,→ ’unknown’)}")
+ # In practice, we would use the editor tool’s apply_fix
,→ command
+ # This is just a demonstration of what could be done
+
+ elif error["type"] == "test_failure":
+ # For test failures, we might suggest reviewing the
,→ implementation
+ resolution_messages.append("Would analyze test failures and
,→ suggest implementation improvements")
+
+ if resolution_messages:
+ return True, "Automated resolution attempted:\n" +
,→ "\n".join(resolution_messages)
+ else:
+ return False, "No automated resolutions available for detected
,→ errors."+
class AgenticSystem:
def __init__(
self,
@@ -243,6 +293,16 @@ Your task is to make changes to the files in the
,→ {self.git_dir} directory to add
safe_log(f"Attempt {attempt + 1} test results: {’PASSED’ if
,→ test_success else ’FAILED’}")
+ # If tests failed, attempt automated error resolution
+ if not test_success:
+ resolved, resolution_message = attempt_error_resolution(
+ self.git_dir, test_output, test_error, self.language
+ )
+ safe_log(f"Error resolution: {resolution_message}")
+
+ # Even if we couldn’t automatically resolve, we still
,→ provide feedback
+ # In a more advanced implementation, we might actually
,→ apply fixes here
+
# If this is the first attempt or tests passed and we
,→ didn’t have a successful attempt yet, update best patch
if attempt == 0 or (test_success and (best_patch is None or
,→ not best_test_results)):
best_patch = current_patch
@@ -278,37 +338,31 @@ Please revise your code to fix these issues and
,→ try again.
# Log final summary
safe_log(f"\n{’=’*20} FINAL SUMMARY {’=’*20}")
safe_log(f"Best solution found on attempt:
,→ {best_test_results[’attempt’] if best_test_results else ’None’}")
- safe_log(f"Tests passed: {best_test_results[’test_success’] if
,→ best_test_results else ’Unknown’}")
+ safe_log(f"Final test result: {’PASSED’ if best_test_results
,→ and best_test_results[’test_success’] else ’FAILED’}")
+
+ if best_test_results:
+ safe_log(f"Final test
,→ output:\n{best_test_results[’test_output’]}")
+ if best_test_results[’test_error’]:
+ safe_log(f"Final test
,→ errors:\n{best_test_results[’test_error’]}")
- # Save attempt history to a file
- history_file =
,→ os.path.join(os.path.dirname(self.chat_history_file),
,→ ’attempt_history.md’)
- with open(history_file, ’w’) as f:
- f.write("# Attempt History\n\n")
- for result in self.attempt_history:
- f.write(f"## Attempt {result[’attempt’]}\n")
- f.write(f"**Tests Passed**: {result[’test_success’]}\n")
- f.write(f"**LLM Calls Used**: {result[’llm_calls’]}\n")
- f.write(f"**Test
,→ Output**:\n‘‘‘\n{result[’test_output’]}\n‘‘‘\n")
- f.write(f"**Test
,→ Error**:\n‘‘‘\n{result[’test_error’]}\n‘‘‘\n")
- f.write(f"**Patch**:\n‘‘‘\n{result[’patch’]}\n‘‘‘\n\n")
+ return bool(best_test_results and
,→ best_test_results[’test_success’])

总结：自改进的现实落点

阅读原文

跳转微信打开

Fish AI Reader

Fish AI Reader

AI辅助创作，多种专业模板，深度分析，高质量内容生成。从观点提取到深度思考，FishAI为您提供全方位的创作支持。新版本引入自定义参数，让您的创作更加个性化和精准。

FishAI

鱼阅，AI 时代的下一个智能信息助手，助你摆脱信息焦虑

联系邮箱 441953276@qq.com