Token导航 LogoToken导航

Agent Skill · 研究检索

conductor-implement(导体工具)

conductor-implement 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。

conductor-implement 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,权限为只读;许可证为 MIT,第三方扫描有未通过项,按安装量在研究检索类的 73,675 个 Skill 里排第 58,594。

总安装 412周安装 17GitHub Stars 692下载 135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
MIT

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:conductor-implement(导体工具)
来源仓库:https://github.com/rmyndharis/antigravity-skills
仓库路径:skills/conductor-implement
安装命令:
npx skills add https://github.com/rmyndharis/antigravity-skills --skill conductor-implement
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/rmyndharis/antigravity-skills --skill conductor-implement

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。

用于执行 track 实现计划中的具体任务,遵循预定义工作流程规则。适合需要最佳实践指导、检查清单或分步验证的开发场景。自动应用相关规范并验证输出结果,确保符合质量标准。仅适用于 implement track 相关任务,其他领域请选用对应技能替代。

conductor-implement 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/conductor-implement

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/rmyndharis/antigravity-skills --skill conducto…;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

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

conductor-implement 怎么安装?

在终端执行 npx skills add https://github.com/rmyndharis/antigravity-skills --skill conductor-implemen…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

conductor-implement 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。

conductor-implement 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

conductor-implement 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

conductor-implement 用的人多吗?

公开数据是累计安装 412 次,近一周安装 17 次,下载 135 次,GitHub 692 星。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。

conductor-implement 安全吗?

第三方扫描有未通过项,风险较高。除非你确认过源码,否则不建议直接安装到常用环境。

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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 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。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 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
conductor-implement

导体工具 · 本页

412692只读MIT未通过

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

Implement Track

Execute tasks from a track's implementation plan, following the workflow rules defined in conductor/workflow.md.

Use this skill when

  • Working on implement track tasks or workflows
  • Needing guidance, best practices, or checklists for implement track

Do not use this skill when

  • The task is unrelated to implement track
  • You need a different domain or tool outside this scope

Instructions

  • Clarify goals, constraints, and required inputs.
  • Apply relevant best practices and validate outcomes.
  • Provide actionable steps and verification.
  • If detailed examples are required, open resources/implementation-playbook.md.

Pre-flight Checks

  1. Verify Conductor is initialized:

- Check conductor/product.md exists - Check conductor/workflow.md exists - Check conductor/tracks.md exists - If missing: Display error and suggest running /conductor:setup first

  1. Load workflow configuration:

- Read conductor/workflow.md - Parse TDD strictness level - Parse commit strategy - Parse verification checkpoint rules

Track Selection

If argument provided:

  • Validate track exists: conductor/tracks/{argument}/plan.md
  • If not found: Search for partial matches, suggest corrections

If no argument:

  1. Read conductor/tracks.md
  2. Parse for incomplete tracks (status [] or [~])
  3. Display selection menu: Select a track to implement: In Progress: 1. [~] auth_20250115 - User Authentication (Phase 2, Task 3) Pending: 2. [] nav-fix_20250114 - Navigation Bug Fix 3. [] dashboard_20250113 - Dashboard Feature Enter number or track ID:

Context Loading

Load all relevant context for implementation:

  1. Track documents:

- conductor/tracks/{trackId}/spec.md - Requirements - conductor/tracks/{trackId}/plan.md - Task list - conductor/tracks/{trackId}/metadata.json - Progress state

  1. Project context:

- conductor/product.md - Product understanding - conductor/tech-stack.md - Technical constraints - conductor/workflow.md - Process rules

  1. Code style (if exists):

- conductor/code_styleguides/{language}.md

Track Status Update

Update track to in-progress:

  1. In conductor/tracks.md:

- Change [] to [~] for this track

  1. In conductor/tracks/{trackId}/metadata.json:

- Set status: "in_progress" - Update updated timestamp

Task Execution Loop

For each incomplete task in plan.md (marked with []):

1. Task Identification

Parse plan.md to find next incomplete task:

  • Look for lines matching - [] Task X.Y: {description}
  • Track current phase from structure

2. Task Start

Mark task as in-progress:

  • Update plan.md: Change [] to [~] for current task
  • Announce: "Starting Task X.Y: {description}"

3. TDD Workflow (if TDD enabled in workflow.md)

Red Phase - Write Failing Test:

Following TDD workflow for Task X.Y...

Step 1: Writing failing test
  • Create test file if needed
  • Write test(s) for the task functionality
  • Run tests to confirm they fail
  • If tests pass unexpectedly: HALT, investigate

Green Phase - Implement:

Step 2: Implementing minimal code to pass test
  • Write minimum code to make test pass
  • Run tests to confirm they pass
  • If tests fail: Debug and fix

Refactor Phase:

Step 3: Refactoring while keeping tests green
  • Clean up code
  • Run tests to ensure still passing

4. Non-TDD Workflow (if TDD not strict)

  • Implement the task directly
  • Run any existing tests
  • Manual verification as needed

5. Task Completion

Commit changes (following commit strategy from workflow.md):

git add -A
git commit -m "{commit_prefix}: {task description} ({trackId})"

Update plan.md:

  • Change [~] to [x] for completed task
  • Commit plan update:
git add conductor/tracks/{trackId}/plan.md
git commit -m "chore: mark task X.Y complete ({trackId})"

Update metadata.json:

  • Increment tasks.completed
  • Update updated timestamp

6. Phase Completion Check

After each task, check if phase is complete:

  • Parse plan.md for phase structure
  • If all tasks in current phase are [x]:

Run phase verification:

Phase {N} complete. Running verification...
  • Execute verification tasks listed for the phase
  • Run full test suite: npm test / pytest / etc.

Report and wait for approval:

Phase {N} Verification Results:
- All phase tasks: Complete
- Tests: {passing/failing}
- Verification: {pass/fail}

Approve to continue to Phase {N+1}?
1. Yes, continue
2. No, there are issues to fix
3. Pause implementation

CRITICAL: Wait for explicit user approval before proceeding to next phase.

Error Handling During Implementation

On Tool Failure

ERROR: {tool} failed with: {error message}

Options:
1. Retry the operation
2. Skip this task and continue
3. Pause implementation
4. Revert current task changes
  • HALT and present options
  • Do NOT automatically continue

On Test Failure

TESTS FAILING after Task X.Y

Failed tests:
- {test name}: {failure reason}

Options:
1. Attempt to fix
2. Rollback task changes
3. Pause for manual intervention

On Git Failure

GIT ERROR: {error message}

This may indicate:
- Uncommitted changes from outside Conductor
- Merge conflicts
- Permission issues

Options:
1. Show git status
2. Attempt to resolve
3. Pause for manual intervention

Track Completion

When all phases and tasks are complete:

1. Final Verification

All tasks complete. Running final verification...
  • Run full test suite
  • Check all acceptance criteria from spec.md
  • Generate verification report

2. Update Track Status

In conductor/tracks.md:

  • Change [~] to [x] for this track
  • Update the "Updated" column

In conductor/tracks/{trackId}/metadata.json:

  • Set status: "complete"
  • Set phases.completed to total
  • Set tasks.completed to total
  • Update updated timestamp

In conductor/tracks/{trackId}/plan.md:

  • Update header status to [x] Complete

3. Documentation Sync Offer

Track complete! Would you like to sync documentation?

This will update:
- conductor/product.md (if new features added)
- conductor/tech-stack.md (if new dependencies added)
- README.md (if applicable)

1. Yes, sync documentation
2. No, skip

4. Cleanup Offer

Track {trackId} is complete.

Cleanup options:
1. Archive - Move to conductor/tracks/_archive/
2. Delete - Remove track directory
3. Keep - Leave as-is

5. Completion Summary

Track Complete: {track title}

Summary:
- Track ID: {trackId}
- Phases completed: {N}/{N}
- Tasks completed: {M}/{M}
- Commits created: {count}
- Tests: All passing

Next steps:
- Run /conductor:status to see project progress
- Run /conductor:new-track for next feature

Progress Tracking

Maintain progress in metadata.json throughout:

{
  "id": "auth_20250115",
  "title": "User Authentication",
  "type": "feature",
  "status": "in_progress",
  "created": "2025-01-15T10:00:00Z",
  "updated": "2025-01-15T14:30:00Z",
  "current_phase": 2,
  "current_task": "2.3",
  "phases": {
    "total": 3,
    "completed": 1
  },
  "tasks": {
    "total": 12,
    "completed": 7
  },
  "commits": [
    "abc1234: feat: add login form (auth_20250115)",
    "def5678: feat: add password validation (auth_20250115)"
  ]
}

Resumption

If implementation is paused and resumed:

  1. Load metadata.json for current state
  2. Find current task from current_task field
  3. Check if task is [~] in plan.md
  4. Ask user: Resuming track: {title} Last task in progress: Task {X.Y}: {description} Options: 1. Continue from where we left off 2. Restart current task 3. Show progress summary first

Critical Rules

  1. NEVER skip verification checkpoints - Always wait for user approval between phases
  2. STOP on any failure - Do not attempt to continue past errors
  3. Follow workflow.md strictly - TDD, commit strategy, and verification rules are mandatory
  4. Keep plan.md updated - Task status must reflect actual progress
  5. Commit frequently - Each task completion should be committed
  6. Track all commits - Record commit hashes in metadata.json for potential revert

平台分布

Claude Code

29.1%
按下载量换算39

Codex

21.65%
按下载量换算29

windsurf

19.24%
按下载量换算26

trae

12.11%
按下载量换算16

Gemini CLI

7.56%
按下载量换算10

Antigravity

3.56%
按下载量换算5

安全审计

Gen Agent Trust Hub

未通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。

继续浏览同类 Skills