← Back to Skills Marketplace
sf0799

Code Orchestrator

by sf0799 · GitHub ↗ · v0.1.0 · MIT-0
cross-platform ✓ Security Clean
95
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install code-orchestrator
Description
Route and sequence coding tasks by selecting and orchestrating code exploration, planning, writing, debugging, refactoring, security, safe commands, and Git...
README (SKILL.md)

编码总控

先判断该调用哪个技能,再决定是否串联多个技能。

如果任务边界不清,先看 routing-examples.md 里的中文示例,再决定路由。

路由规则

  • 需要了解项目结构、入口、依赖、调用链、数据流时,调用 $code-explore
  • 需要先出方案、拆步骤、评估风险和实现顺序时,调用 $code-architect
  • 需要直接写代码、补代码、改逻辑、补测试时,调用 $code-write
  • 需要定位报错、分析根因、给出最小修复时,调用 $code-debug
  • 需要在不改行为前提下清理结构和重复逻辑时,调用 $code-refactor
  • 需要检查漏洞、权限、输入校验、密钥泄露等问题时,调用 $code-security
  • 需要执行构建、测试、lint、format、安装依赖等安全命令时,调用 $shell-safe-exec
  • 需要处理分支、提交、历史保护等 Git 事项时,调用 $git-discipline

编排原则

  1. 不要默认把所有技能都用一遍,只选择当前任务真正需要的最小集合。
  2. 如果任务边界不清,先探索再决策,不要直接写代码。
  3. 如果任务跨多个阶段,按“分析 -> 规划 -> 实现 -> 验证 -> 提交”顺序串联。
  4. 如果用户只要单一步骤,就只调用对应技能,不强行增加流程。
  5. 如果风险高或影响面大,优先加上验证和安全检查环节。

常见流程

  • 新功能: $code-explore -> $code-architect -> $code-write -> $shell-safe-exec
  • 修 bug: $code-explore -> $code-debug -> $code-write -> $shell-safe-exec
  • 纯重构: $code-explore -> $code-refactor -> $shell-safe-exec
  • 安全修复: $code-explore -> $code-security -> $code-write -> $shell-safe-exec
  • 发布前整理: $code-explore -> $git-discipline

输出

  • 选中的技能
  • 调用顺序
  • 每个技能被选中的原因
  • 本轮应执行到哪一步为止
Usage Guidance
This orchestrator is coherent and low-risk by itself because it only chooses and sequences other skills and has no installs or credential requests. Before installing, review the SKILL.md and manifest of each downstream skill it calls (especially $shell-safe-exec and $git-discipline) to confirm they don't request excessive credentials, run arbitrary network downloads, or execute unsafe host-level commands. Also verify the downstream skills' install specs and environment requirements; the orchestrator inherits their risk. The agent YAML's allow_implicit_invocation: false is a helpful safeguard — keep it if you want to avoid implicit activation.
Capability Analysis
Type: OpenClaw Skill Name: code-orchestrator Version: 0.1.0 The 'code-orchestrator' skill bundle serves as a high-level routing logic for an AI agent to manage coding tasks. The instructions in SKILL.md and routing-examples.md are focused on task decomposition and SDLC workflows (analysis, planning, implementation, and security review) without any evidence of malicious intent, data exfiltration, or unauthorized command execution.
Capability Assessment
Purpose & Capability
Name/description match the SKILL.md: the file's sole job is to choose and sequence other coding skills (explore, architect, write, debug, refactor, security, shell-safe-exec, git-discipline). It does not request unrelated credentials, binaries, or installs.
Instruction Scope
The instructions are limited to deciding which skills to call, in what order, and why. This orchestration delegates potentially sensitive actions (running commands, modifying repo, checking secrets) to the referenced skills (e.g., $shell-safe-exec, $git-discipline). The orchestrator itself does not instruct reading arbitrary host files or contacting external endpoints, but its safety depends on the invoked skills' scopes and permissions.
Install Mechanism
Instruction-only skill with no install spec and no code files that execute — low risk from installation. Included files are a routing reference and a small agent YAML, which do not contain executable install instructions.
Credentials
The skill declares no required environment variables or credentials. Note: actual environment/credential needs will come from the downstream skills it routes to, so those should be reviewed separately.
Persistence & Privilege
always is false and the agent config explicitly sets allow_implicit_invocation: false, so it will not be automatically invoked implicitly. It does not request persistent system presence or modify other skills' configs.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install code-orchestrator
  3. After installation, invoke the skill by name or use /code-orchestrator
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
initial release
Metadata
Slug code-orchestrator
Version 0.1.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Code Orchestrator?

Route and sequence coding tasks by selecting and orchestrating code exploration, planning, writing, debugging, refactoring, security, safe commands, and Git... It is an AI Agent Skill for Claude Code / OpenClaw, with 95 downloads so far.

How do I install Code Orchestrator?

Run "/install code-orchestrator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Code Orchestrator free?

Yes, Code Orchestrator is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Code Orchestrator support?

Code Orchestrator is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Code Orchestrator?

It is built and maintained by sf0799 (@sf0799); the current version is v0.1.0.

💬 Comments