Skip to content

anbeime/skill / skills/ledger-tasks-yylo

ledger-tasks-yylo

Verified

Drive multi-step coding-agent work through a git-native Kanban/task ledger. Use when an agent needs persistent task state across sessions, dependency-aware ordering, status transitions with receipts, or serialized multi-worktree delivery. For task/board operations, not memory or general notes.

FileKindSize
README.mdmarkdown1.5 KB
SKILL.mdskill md2.8 KB
metadata.jsonconfig718 B
# YYLO Ledger Tasks (ledger-tasks-yylo)

让 AI 编程代理通过 git 原生的看板/任务账本驱动多步骤开发:任务状态、依赖排序、状态回执与多 worktree 串行交付,全部以文件形式保存在仓库内,跨会话可审计。

## 适用场景

- 多步骤、跨会话的功能开发与任务追踪
- 依赖感知的并行执行排序(拓扑排序、阻塞关系)
- 状态流转必须留痕(`--response` 回执 + `--commit` 关联)
- 多 agent / 多 worktree 协作的串行合并交付

## 核心原则

1. 先读后写:`yy ledger list` / `get` / `ready` 了解现状再变更。
2. 任务粒度小:一次迭代可完成,避免撑爆上下文。
3. 每次状态变更必须带 `--response` 回执,说明做了什么、如何验证。
4. 依赖显式建模:`deps add` 声明阻塞关系,环会被自动检测。
5. 绝不直接编辑 `.juno_task/` 文件,CLI 是唯一合法入口。

## 安装

```bash
npm install -g @yylo/cli
```

要求 Node >= 20。

## 常用命令

```bash
yy ledger create "任务描述" --status todo --tags feature
yy ledger ready                 # 查看无阻塞任务
yy ledger order --scores        # 拓扑排序,规划并行执行
yy ledger mark in_progress --id TASK_ID --response "开始实现"
yy ledger mark done --id TASK_ID --response "完成并测试" --commit abc123
```

## 技能来源

来源:<https://github.com/yylo-dev/yylo-skills>(MIT)。
依赖外部 YYLO CLI:<https://github.com/yylo-dev/yylo>。
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall