跳到正文
TakoAPI
返回技能列表

autonomous-executor

由 ishsharm0 提供

Productivity & Tasks
—暂无评分
你的评分

简介

Self-healing, error-recovering task execution with full autonomy.

描述

该技能的功能

A Python execution wrapper that automatically categorizes errors and applies recovery strategies, including retries with exponential backoff, dependency installation, and resource creation. Supports checkpoint/resume for long-running jobs and overnight batch queuing. Logs all execution history to MongoDB.

Eliminates boilerplate retry and error-handling code by wrapping any function with categorized recovery logic out of the box.

何时使用

  • Queue overnight batch deployments to run while sleeping
  • Retry flaky API calls without writing custom retry logic
  • Resume a long task from a checkpoint after a failure
  • Auto-install missing Python packages mid-execution
  • Review morning reports on overnight task outcomes

示例工作流

输入

User asks: deploy my portfolio project overnight and report results in the morning

智能体
  1. 1

    Import Overnight from autonomous_executor.overnight

  2. 2

    Queue the deploy task with Overnight.deploy('my-portfolio', 'portfolio')

  3. 3

    Call Overnight.start() to begin scheduled execution

  4. 4

    Executor retries on network or rate-limit failures using exponential backoff

  5. 5

    Checkpoints progress so execution can resume if interrupted

输出

Morning report showing task completion status, retry attempts, duration, and any errors encountered

前置要求

  • MongoDB instance with connection credentials

安装

clawhub install ishsharm0/autonomous-executor

运行 CLI 命令,或把这段提示词粘贴到 Claude Code、Cursor、Codex 或你在用的任意编程智能体里。

链接