140 lines
6.0 KiB
Python
140 lines
6.0 KiB
Python
import os
|
||
import re
|
||
import yaml
|
||
import logging
|
||
import subprocess
|
||
import hashlib
|
||
from pathlib import Path
|
||
from datetime import datetime
|
||
from typing import List, Dict, Any
|
||
|
||
# 配置审计日志
|
||
logging.basicConfig(
|
||
filename='skill_audit.log',
|
||
level=logging.INFO,
|
||
format='%(asctime)s [%(levelname)s] %(message)s'
|
||
)
|
||
|
||
class IndustrialSkillEngine:
|
||
def __init__(self, skills_dir: str, llm_handler):
|
||
self.root = Path(skills_dir).resolve()
|
||
self.llm = llm_handler
|
||
self.registry = {}
|
||
self.session_id = hashlib.md5(str(datetime.now()).encode()).hexdigest()[:8]
|
||
# 状态机：记录当前任务执行到的步骤
|
||
self.state = {"current_skill": None, "history": [], "pending_params": []}
|
||
|
||
# --- 1. 工业级初始化：依赖检查与索引 ---
|
||
def boot(self):
|
||
for skill_md in self.root.glob("**/SKILL.md"):
|
||
with open(skill_md, 'r', encoding='utf-8') as f:
|
||
content = f.read()
|
||
meta = yaml.safe_load(re.search(r'^---(.*?)---', content, re.DOTALL).group(1))
|
||
name = meta.get('name')
|
||
|
||
# 检查依赖 (Pre-flight check)
|
||
req_file = skill_md.parent / "requirements.txt"
|
||
has_deps = req_file.exists()
|
||
|
||
self.registry[name] = {
|
||
"root": skill_md.parent,
|
||
"meta": meta,
|
||
"content": content,
|
||
"has_deps": has_deps
|
||
}
|
||
logging.info(f"Engine Booted. Session: {self.session_id}. Skills: {list(self.registry.keys())}")
|
||
|
||
# --- 2. 自动化依赖环境隔离 (venv 思想) ---
|
||
def _ensure_dependencies(self, skill_name: str):
|
||
skill = self.registry[skill_name]
|
||
if skill["has_deps"]:
|
||
# 工业级引擎通常会检查一个隐藏的 .installed 标识
|
||
if not (skill["root"] / ".deps_installed").exists():
|
||
print(f"📦 正在为技能 {skill_name} 安装必要依赖...")
|
||
subprocess.run(["pip", "install", "-r", "requirements.txt"], cwd=skill["root"])
|
||
(skill["root"] / ".deps_installed").touch()
|
||
|
||
# --- 3. 增强版安全执行器：带重试逻辑与审计 ---
|
||
def _execute_with_retry(self, command: str, skill_name: str, retry_count=1) -> str:
|
||
self._ensure_dependencies(skill_name)
|
||
|
||
# 安全预检：禁止敏感命令
|
||
forbidden = ["rm ", "> /dev/", "chmod", "sudo"]
|
||
if any(f in command for f in forbidden):
|
||
return "🚫 安全风险：检测到非法指令，执行被拦截。"
|
||
|
||
# 权限确认 (Y/N)
|
||
print(f"\n\033[1;33m[Audit ID: {self.session_id}]\033[0m 请求执行: {command}")
|
||
confirm = input("是否授权执行？(y/n/skip): ").lower()
|
||
if confirm != 'y': return "Execution skipped by user."
|
||
|
||
logging.info(f"Executing: {command} in {skill_name}")
|
||
|
||
try:
|
||
env = os.environ.copy()
|
||
env["SKILL_CONTEXT"] = skill_name
|
||
|
||
res = subprocess.run(command, shell=True, cwd=self.registry[skill_name]["root"],
|
||
env=env, capture_output=True, text=True, timeout=30)
|
||
|
||
if res.return_code != 0:
|
||
# 工业级特性：自动将错误回传给 LLM 进行自愈 (Self-healing)
|
||
logging.error(f"Command failed: {res.stderr}")
|
||
if retry_count > 0:
|
||
print(f"⚠️ 执行失败，尝试让 AI 自愈修复参数...")
|
||
new_prompt = f"命令 '{command}' 失败，错误信息: {res.stderr}。请根据错误重新生成正确的命令，或提示用户补全参数。"
|
||
# 这里会递归调用逻辑进行修复
|
||
return self.run(new_prompt, is_retry=True)
|
||
return f"Error: {res.stderr}"
|
||
|
||
return res.stdout
|
||
except Exception as e:
|
||
return str(e)
|
||
|
||
# --- 4. 递归文档注入与意图路由 ---
|
||
def _get_expanded_context(self, skill_name: str, user_prompt: str):
|
||
skill = self.registry[skill_name]
|
||
base_content = skill["content"]
|
||
|
||
# 扫描目录下的辅助文件夹
|
||
sub_dirs = ["reference", "examples"]
|
||
found_docs = []
|
||
for d in sub_dirs:
|
||
path = skill["root"] / d
|
||
if path.exists():
|
||
found_docs.extend([f"{d}/{f.name}" for f in path.glob("*.md")])
|
||
|
||
if found_docs:
|
||
# 仅在模型认为有必要时，“点餐式”加载
|
||
choice = self.llm(f"用户问题: {user_prompt}\n可选深入文档: {found_docs}\n需要读取哪个？(仅返回路径，不需则返回 None)")
|
||
if choice != "None" and any(choice in doc for doc in found_docs):
|
||
with open(skill["root"] / choice, 'r') as f:
|
||
print(f"📂 深度加载: {choice}")
|
||
base_content += f"\n\n--- 深度参考 ({choice}) ---\n{f.read()}"
|
||
|
||
return base_content
|
||
|
||
# --- 5. 主运行接口 ---
|
||
def run(self, user_prompt: str, is_retry=False):
|
||
# 如果是重试，跳过技能选择
|
||
if not is_retry:
|
||
skill_map = {n: v["description"] for n, v in self.registry.items()}
|
||
target = self.llm(f"用户意图: {user_prompt}\n可选技能清单: {skill_map}\n请返回匹配的技能名:")
|
||
self.state["current_skill"] = target
|
||
|
||
skill_name = self.state["current_skill"]
|
||
if skill_name not in self.registry: return "Skill not found."
|
||
|
||
# 获取递归上下文
|
||
context = self._get_expanded_context(skill_name, user_prompt)
|
||
|
||
# 决策：是直接回答还是执行脚本
|
||
decision = self.llm(f"上下文: {context}\n问题: {user_prompt}\n决定动作：EXEC: <command> 或 ANSWER: <text>")
|
||
|
||
if "EXEC:" in decision:
|
||
cmd = decision.split("EXEC:")[1].strip()
|
||
return self._execute_with_retry(cmd, skill_name)
|
||
|
||
return decision.replace("ANSWER:", "").strip()
|
||
|