← 返回 Skills 市场

Kai Master Builder

Name: Kai Master Builder
Author: ogdegenblaze

作者 Blaze🔥 · GitHub ↗ · v1.0.1 · MIT-0

cross-platform ⚠ suspicious

159

总下载

当前安装

版本数

在 OpenClaw 中安装

/install kai-master-builder

功能描述

Developer, Code, Engineer. Guide the Agent to build apps/features/goals efficiently and securely. Creates project plans, task lists, and provides building pr...

使用说明 (SKILL.md)

Kai Master Builder

Build anything efficiently with structured plans and automated execution.

When to Use

User wants to build a new app/feature/fix
Need structured approach to complex project
Want automated building via cron
Need to track progress across sessions

The Process

Phase 1: Create Project Plan

Input: GOAL (what to build/improve/fix)

Output: PROJECT/PROJECT_PLAN.md

# Project Plan: [PROJECT_NAME]

## Overview
What this project does and why.

## Goals
- Primary goal
- Secondary goals

## Architecture/Approach
How it will be built.

## Security Checklist
- [ ] No API keys hardcoded (use env vars)
- [ ] No .env loading in scripts
- [ ] All secrets in config, not code
- [ ] Input validation
- [ ] Error handling

## Risks & Mitigations
| Risk | Impact | Mitigation |
|------|--------|------------|
| ... | ... | ... |

## Dependencies
- External: ...
- Internal: ...

## Milestones
1. [ ] Phase 1: ...
2. [ ] Phase 2: ...
3. [ ] Phase 3: ...

## Validation
How to test/validate each milestone.

Phase 2: Create Task List

Output: PROJECT/PROJECT_TASKS.md

# Tasks for: [PROJECT_NAME]

## Task Format
- [ ] TASK_NAME | status: TODO | depends: none | validated_by: test command

## Tasks

### Phase 1: Setup
- [ ] Create project structure | status: TODO | depends: none | validated_by: ls project/
- [ ] Setup dependencies | status: TODO | depends: 1 | validated_by: pip list / npm list

### Phase 2: Core Feature
- [ ] Implement X | status: TODO | depends: 1 | validated_by: python3 test_x.py

### Phase 3: Testing & Polish
- [ ] Write tests | status: TODO | depends: 2 | validated_by: pytest
- [ ] Update docs | status: TODO | depends: 2 | validated_by: doc builds

## Blocked
- Task X: waiting on external API access

## Done
(Completed tasks move here with timestamp)

Phase 3: Build!

Prompt provided to Agent/User:

BUILDER PROMPT

You are building: [PROJECT_NAME]

Context:

Read PROJECT/PROJECT_PLAN.md for architecture and goals
Read PROJECT/PROJECT_TASKS.md for task list
Follow the iteration loop below

Iteration Loop:

Pick a task - Choose TODO task whose dependencies are all DONE
Implement it - Write the code
Test it - Run validation command
Document - Update CHANGELOG.md with what you built
Mark DONE - Update TASKS.md, move to Done section with timestamp
Report - Report back to main session

If blocked:

Mark task as BLOCKED with reason
Report what you need to proceed

When all tasks done:

Report: "PROJECT is DONE! Built with kai-master-builder."
Update PROJECT_README.md with final state

Automation

Setup cron for autonomous building:

/call_skill kai-master-builder --build PROJECT_NAME

Or let Agent run it via sessions_spawn with this prompt.

Project Folder Structure

PROJECT/
├── PROJECT_PLAN.md      # Architecture and goals
├── PROJECT_TASKS.md     # Task list with status
├── PROJECT_CHANGELOG.md # What was built
├── PROJECT_README.md    # Documentation
└── src/                 # Source code

Tips

Break big tasks into \x3C 1 hour chunks
Every task should have clear validation
Mark tasks BLOCKED if waiting on something
Update CHANGELOG after each task
Backup TASKS.md before modifying

kai-master-builder: Turn goals into working code

安全使用建议

This skill appears to do what it says (plan tasks and run build/test steps), but exercise caution before using it unattended. Things to consider before installing or scheduling it: 1) Run it in an isolated workspace or container (so tests/builds can't touch your system or secrets). 2) Review any generated/modified code and commits before pushing them. 3) Ensure you understand and restrict what validation commands will run (pytest, npm, pip, doc builds) and install only the tools you trust. 4) Do not grant access to sensitive environment variables or credentials; the skill doesn't need any declared secrets, but build scripts might. 5) Because provenance is unknown (no homepage/source), prefer running it interactively first and require manual approval before enabling cron/automation. If the owner/source is provided and the SKILL.md is updated to list all required binaries and explicit safety guardrails, my confidence would increase.

功能分析

Type: OpenClaw Skill Name: kai-master-builder Version: 1.0.1 The skill is a project management framework designed to help an AI agent build software projects through structured planning and task tracking. It includes templates for project plans and task lists, and even promotes security best practices like input validation and avoiding hardcoded secrets in its 'Security Checklist' section (SKILL.md).

能力评估

ℹ Purpose & Capability

The skill is an instruction-only 'builder' that creates plans, tasks, and run/test prompts — python3 is a reasonable required binary. However, the runtime instructions reference commands/tools not declared as required (e.g., npm, pytest, pip, doc build commands). The presence of those references without declaring them is an inconsistency (minor but notable).

⚠ Instruction Scope

The SKILL.md tells the agent to pick tasks, implement code, run validation commands (python3 test_x.py, pytest, pip list, npm list, doc builds) and update project files. Those actions will execute code and modify files in the user's workspace. While expected for a builder skill, the instructions also explicitly encourage autonomous operation (cron / sessions_spawn). Executing arbitrary build/test commands can access the network, installed packages, or system resources — the skill gives broad discretion to run such commands without guardrails.

✓ Install Mechanism

No install spec or code is included (instruction-only). That minimizes supply-chain risk because nothing is downloaded or written by an installer.

ℹ Credentials

The skill requests no environment variables or credentials, and its security checklist discourages hardcoding secrets. That is proportionate. Still, because it runs arbitrary build/test commands, it could read local environment or files if the agent is instructed to do so — the SKILL.md does not enumerate or limit which environment variables or paths the agent may access.

⚠ Persistence & Privilege

The skill itself is not always-enabled (always: false) and has no install persistence. However, the SKILL.md explicitly recommends setting up cron or using sessions_spawn to run autonomously. Scheduling this skill to run unattended increases risk because it may execute build/test steps and modify files without an explicit human review.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install kai-master-builder
安装完成后，直接呼叫该 Skill 的名称或使用 /kai-master-builder 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.1

- Added keywords "Developer, Code, Engineer" to the skill description for improved discoverability. - No functional or process changes; documentation only.

v1.0.0

kai-master-builder 1.0.0 - Initial release of kai-master-builder. - Generates secure, structured project plans and detailed task lists for building apps or features. - Guides users (or runs autonomously) through an iterative build/test/report/document loop. - Supports automated execution via cron or user-driven prompts. - Encourages strong security practices and clear validation for every task. - Tracks progress and handles blocked tasks for efficient project management.

元数据

Slug kai-master-builder

版本 1.0.1

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 2

常见问题