Token导航 LogoToken导航

Agent Skill · 研究检索

executing-subtask(执行子任务)

executing-subtask 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

executing-subtask 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为调用外部服务;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 63,254。

总安装 318周安装 13GitHub Stars 公开资料未说明下载 103核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:executing-subtask(执行子任务)
来源仓库:https://github.com/b-mendoza/agent-skills
仓库路径:skills/executing-subtask
安装命令:
npx skills add https://github.com/b-mendoza/agent-skills --skill executing-subtask
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/b-mendoza/agent-skills --skill executing-subtask

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

执行子任务技能用于将复杂任务拆解为单一可执行单元。适合在保持主上下文清晰的前提下委派具体实现工作。需配合工单快照和任务计划文档使用,确保输入完整。使用前应验证所需文档是否存在,避免因依赖缺失导致失败。executing-subtask 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/executing-subtask

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/b-mendoza/agent-skills --skill executing-subta…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要调用外部服务,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

executing-subtask 怎么安装?

在终端执行 npx skills add https://github.com/b-mendoza/agent-skills --skill executing-subtask。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

executing-subtask 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

executing-subtask 需要什么权限?

权限级别是调用外部服务。该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

executing-subtask 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

executing-subtask 用的人多吗?

公开数据是累计安装 318 次,近一周安装 13 次,下载 103 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

executing-subtask 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
executing-subtask

执行子任务 · 本页

318—调用外部服务未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Executing Subtask

Purpose

Execute exactly ONE task from the task plan. Delegate the actual implementation to the task-executor subagent to keep the main agent's context clean for coordination and review. After execution, validate the work and update tracking artifacts.

Inputs

InputSourceRequiredExample
TICKET_KEYUser / $ARGUMENTSYesJNS-6065
TASK_NUMBERUser / $ARGUMENTSYes3

Both the ticket snapshot (docs/<TICKET_KEY>.md) and the task plan (docs/<TICKET_KEY>-tasks.md) must exist. If either is missing, tell the user which prerequisite skill to run.

Input contract (produced by upstream skills)

The task plan file docs/<TICKET_KEY>-tasks.md must contain these sections, built up across the preceding four phases:

Required section / elementProduced byUsed in stepWhy
## Task <N>: with all 8 subsectionsplanning-jira-tasksStep 1 (load)Source content for the execution brief
## Dependency Graphplanning-jira-tasksStep 1 (pre-flight)Validates dependencies are satisfied
## Decisions Logclarifying-assumptionsStep 2 (brief)Resolved decisions folded into execution context
Per-task Questions to answer resolvedclarifying-assumptionsStep 1 (pre-flight)Pre-flight checks all questions are answered
## Jira Subtasks table with keyscreating-jira-subtasksStep 5b (Jira)Maps task number to Jira subtask key for status
Jira Subtask: <KEY> in each task sectioncreating-jira-subtasksStep 5b (Jira)Identifies which Jira issue to transition
**Status:** on previously completed tasksexecuting-subtask (self)Step 1 (pre-flight)Checks whether dependencies are marked complete

Pre-flight gate: If the ## Jira Subtasks table is missing, subtasks were not created in Jira. Warn the user and ask whether to proceed without Jira integration or run the creating-jira-subtasks skill first. This is a warning, not a hard block — execution can proceed without Jira tracking.

Subagent Registry

SubagentPathPurpose
task-executor./subagents/task-executor.mdPerforms the actual implementation

Before delegating, read the subagent file to understand its contract (expected input format, output format, and rules). The path is relative to this skill's directory.

Output

  • Implemented code / configuration changes for the specified task.
  • Updated task plan with execution status.
  • Jira subtask transitioned (if Jira MCP available and subtask keys present).

Output contract (consumed by orchestrator and self)

After this skill completes for a given task, the plan file must contain these updates:

AdditionConsumed byWhy
**Status:** ✅ Complete (<date>) on taskorchestrator, selfOrchestrator tracks progress; self checks deps
**Implementation summary:** on taskorchestratorProgress file gets a concise summary
**Files changed:** list on taskorchestratorProgress reporting
## Jira Subtasks table status updated to DoneorchestratorReflects current state

Execution Steps

1. Load and validate the task

Read docs/<TICKET_KEY>-tasks.md and extract ## Task <TASK_NUMBER>.

Pre-flight checks — stop if any fail:

  • The task exists in the plan.
  • Dependencies listed in Dependencies / prerequisites are marked complete (look for **Status:** ✅ Complete on each dependency's task section).
  • Questions to answer before starting are all resolved (no unresolved items without a recorded fallback — look for strikethrough + answer format or None).

If pre-flight fails, tell the user what needs to be resolved first and stop.

2. Prepare the execution brief

Build a self-contained execution brief that includes ONLY what the subagent needs:

# Execution Brief — <TICKET_KEY> Task <N>: <Title>

## Objective

<from task plan>

## Relevant Requirements and Context

<from task plan, plus any resolved decisions from the Decisions Log>

## Implementation Notes

<from task plan — must reflect any updates applied during clarification>

## Definition of Done

<from task plan>

## Likely Files / Artifacts Affected

<from task plan>

## Resolved Questions & Decisions

<any answers from the Decisions Log that affect this task>

## Constraints

- Only implement what is described above.
- Do not modify files unrelated to this task.
- If you encounter ambiguity not covered here, STOP and report it — do not guess.
- Run existing tests to verify you haven't broken anything.
- If the definition of done includes new tests, write them.

Write this brief to docs/<TICKET_KEY>-task-<N>-brief.md.

3. Delegate to the task-executor subagent

agent task-executor "Execute the task described in docs/<TICKET_KEY>-task-<N>-brief.md"

The subagent performs the implementation in an isolated context.

4. Review the subagent's work

After the subagent completes, review:

  • Definition of done: Check each condition against actual changes.
  • Scope containment: Verify no files outside Likely files affected were modified unexpectedly (unless justified).
  • Tests pass: Run the project's test suite (or relevant subset).
  • No regressions: If existing tests exist, confirm they still pass.

If review finds issues, provide specific feedback and re-invoke the subagent with corrections. Limit to 3 retry cycles — after that, report remaining issues to the user.

5. Update tracking

a. Update the task plan

In docs/<TICKET_KEY>-tasks.md, update the task section:

**Status:** ✅ Complete (<YYYY-MM-DD>)
**Implementation summary:** <2–3 sentence summary of what was done>
**Files changed:**

- `path/to/file1.ts` — <what changed>
- `path/to/file2.ts` — <what changed>
b. Update Jira (if MCP available)

Look up the Jira subtask key from the Jira Subtask: <KEY> line in the task section, or from the ## Jira Subtasks table.

  • Transition the subtask to "In Progress" at the start of execution (Step 3).
  • Transition the subtask to "Done" after successful review.
  • Add a comment to the subtask summarizing what was implemented.

If the Jira subtask key is not present (subtasks were never created), skip Jira updates silently — do not error.

c. Update the Jira Subtasks table

If the ## Jira Subtasks table exists, update the Status column for this task from To Do to Done.

d. Clean up

Delete the temporary execution brief file: docs/<TICKET_KEY>-task-<N>-brief.md

After deletion, verify the file no longer exists. If deletion fails (e.g., permission error), log a warning but do not block — cleanup failure is non-critical.

6. Report to user

Task <N> complete: <Title>

Summary: <what was done in 2–3 sentences>

Files changed:
- <list>

Tests: <passing / N new tests added>

Jira subtask <KEY>: transitioned to Done.

⚠️ Remaining items (if any):
- <anything that couldn't be fully resolved>

Ready for the next task? Let me know which one to tackle.

Safety Rules

  • One task at a time. Never auto-continue to the next task.
  • Scope discipline. Do not implement anything outside the task's scope, even if it seems like a quick win.
  • Fail loudly. If the subagent encounters ambiguity or a blocker, surface it to the user immediately rather than making assumptions.
  • Preserve the plan. The task plan is the source of truth. If execution reveals the plan needs changes, propose the change to the user — don't silently modify the plan.

平台分布

Codex

38.01%
按下载量换算39

Claude

28.93%
按下载量换算30

Cursor

19.03%
按下载量换算20

Gemini CLI

8.76%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

Snyk

通过

权限和风险

调用外部服务

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills