Three-route soft router for pressured coding-agent work.
Security
编码总控
Try itRoute and sequence coding tasks by selecting and orchestrating code exploration, planning, writing, debugging, refactoring, security, safe commands, and Git...
What it does
Route and sequence coding tasks by selecting and orchestrating code exploration, planning, writing, debugging, refactoring, security, safe commands, and Git...
The skill document
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
- 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。
- 如果任务边界不清,先探索再决策,不要直接写代码。
- 如果任务跨多个阶段,按“分析 -> 规划 -> 实现 -> 验证 -> 提交”顺序串联。
- 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。
- 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
Related skills
AI coding agent powered by CellCog Co-work. Code generation, debugging, refactoring, codebase exploration, terminal operations — executed directly on your machine. Lightweight with multimedia tools loaded on demand.
Build a compact local SQLite index of every function, class, method, interface and type across your repos, so an agent finds a symbol's file:line and signatu...
Delegate coding, repository analysis, file edits, test runs, or code review to the local Codex CLI without embedding an OpenAI API key. This skill was created to work with ChatGPT/Codex enterprise accounts that may not have the same level of API access. Invokes codex exec with existing ChatGPT/Codex CLI authentication and returns Codex's final output.
Repo-aware router skill for AI coding CLIs: map tasks to the right repository, project skill or agent, and native CLI backend across OpenClaw, Claude Code, O...
Use this skill when OpenClaw should hand a coding task to OpenCode, keep the OpenCode run model-compatible with Z.AI Coding Plan / GLM plans, and mirror exec...