ai-agent-book 精选快照(<2MB 代码与文档,来自 github.com/bojieli/ai-agent-book)
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
This commit is contained in:
@@ -0,0 +1,73 @@
|
||||
"""子进程 Python 沙箱:在隔离的子进程中执行模型生成的代码,带超时限制。
|
||||
|
||||
预装 sympy / numpy / scipy(随本项目 requirements 一并安装)。
|
||||
之所以用子进程而不是 exec(),是为了:
|
||||
1. 隔离:模型代码崩溃 / 死循环不会影响主进程;
|
||||
2. 超时:可强制杀掉执行过久的代码;
|
||||
3. 干净的命名空间:每次执行都是全新解释器。
|
||||
"""
|
||||
|
||||
import subprocess
|
||||
import sys
|
||||
import tempfile
|
||||
import os
|
||||
|
||||
# 在执行前注入的前置代码:预导入常用数学库,方便模型直接使用。
|
||||
PREAMBLE = """
|
||||
import math
|
||||
import sympy
|
||||
import numpy as np
|
||||
import scipy
|
||||
from sympy import *
|
||||
"""
|
||||
|
||||
|
||||
def run_python(code: str, timeout: int = 20) -> str:
|
||||
"""在子进程沙箱中执行 Python 代码,返回 stdout/stderr 文本。
|
||||
|
||||
参数:
|
||||
code: 待执行的 Python 源码(模型生成)。
|
||||
timeout: 超时秒数,超过则终止子进程。
|
||||
返回:
|
||||
执行输出(合并 stdout 与 stderr)。模型应通过 print() 输出结果。
|
||||
"""
|
||||
full_code = PREAMBLE + "\n" + code
|
||||
|
||||
# 写入临时文件,用当前解释器(已装好 sympy/numpy/scipy)执行。
|
||||
with tempfile.NamedTemporaryFile(
|
||||
mode="w", suffix=".py", delete=False, encoding="utf-8"
|
||||
) as f:
|
||||
f.write(full_code)
|
||||
path = f.name
|
||||
|
||||
try:
|
||||
proc = subprocess.run(
|
||||
[sys.executable, path],
|
||||
capture_output=True,
|
||||
text=True,
|
||||
timeout=timeout,
|
||||
# 在临时目录里执行,避免误读写项目文件(仍能访问已安装的库)。
|
||||
cwd=tempfile.gettempdir(),
|
||||
)
|
||||
out = proc.stdout
|
||||
if proc.stderr.strip():
|
||||
out += "\n[stderr]\n" + proc.stderr
|
||||
if not out.strip():
|
||||
out = "(代码没有任何输出,请记得用 print() 打印结果)"
|
||||
# 截断超长输出,避免撑爆上下文。
|
||||
return out[:4000]
|
||||
except subprocess.TimeoutExpired:
|
||||
return f"[错误] 代码执行超时(超过 {timeout} 秒),可能存在死循环或计算量过大。"
|
||||
except Exception as e: # noqa: BLE001
|
||||
return f"[错误] 执行失败: {e}"
|
||||
finally:
|
||||
try:
|
||||
os.unlink(path)
|
||||
except OSError:
|
||||
pass
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
# 简单自测
|
||||
print(run_python("from sympy import factorint; print(factorint(360))"))
|
||||
print(run_python("print(sum(1 for n in range(1,2025) if n%3 and n%5 and n%7))"))
|
||||
Reference in New Issue
Block a user