工作流能力诊断

sodejm2026-10-25查看 774复制 294

检查主机平台、工具前提条件、包结构以及能力真实性宣传诊断。

📘如何使用这个技能

1
复制技能内容·点击上方「Copy SKILL.md」复制全部内容，或 Download 下载到本地
2
打开你的 AI 客户端·支持 Claude Code、Trae、Cursor、ChatGPT、DeepSeek、Kimi 等任意 AI 助手
3
粘贴为系统提示/知识文件·Claude Code：放入 .claude/skills/ 目录；普通对话：作为首条消息发送并说「请严格按这份技能文档执行」
4
直接提出你的需求·AI 会按技能定义的工作流、格式和约束来处理，无需再解释背景

💡 提示：技能生效后，每次对话只需一句需求，AI 会按 SKILL.md 的流程执行。

SKILL.md2212 chars

---
name: workflow-capability-diagnostics
description: Inspect host platforms, tool prerequisites, package structures, and capability truth-in-advertising diagnostics.
---

# Workflow Capability Diagnostics

Run comprehensive system platform checks, host tool prerequisite inspections against tested matrices, package manifest/skills/playbook/validator structure verification, and capability truth-in-advertising audits across all registered packages.

## Core Responsibilities

1. **System Platform Verification**:
   - Detect host operating system, architecture, and Python runtime version.
   - Verify compatibility against the tested OS baseline (`linux`, `darwin`).
2. **Tool Prerequisite Diagnostics**:
   - Query presence and versions of standard and domain security tools (e.g., `python3`, `cops`, `git`, `nmap`, `trivy`, `curl`, `jq`, `docker`).
   - Distinguish missing tools explicitly without treating them as silent failures in offline planning mode.
3. **Package Structure Diagnostics**:
   - Verify package manifest (`package.json`, `plugin.json`), documentation (`README.md`, `PLAYBOOK.md`), validation scripts (`validate-package.py`), and skills directories.
   - Report status as `ready`, `degraded`, or `unready`.
4. **Capability Truth Audit**:
   - Ensure zero unverified claims of `live-validated` mode.
   - Reconcile plugins, specialist profiles, and scenarios against catalog definitions.

## CLI Usage

### Run All Diagnostics

```bash
python3 -m cops diagnostics
```

### Inspect Host Tools Only

```bash
python3 -m cops diagnostics --tools
```

### Scope to a Specific Package

```bash
python3 -m cops diagnostics --package offensive-engagement-workbench
```

### Strict Mode Gate

Fails (exit code 1) if any package is degraded/unready or any tool is missing:

```bash
python3 -m cops diagnostics --strict
```

### JSON Output for Automation

```bash
python3 -m cops diagnostics --json
```

## Python API

```python
from pathlib import Path
from cops.diagnostics import run_diagnostics

report = run_diagnostics(root=Path("."), strict=False)
assert report.system.is_supported
assert report.capability_truth_passed
for pkg in report.packages:
    print(pkg.package_id, pkg.status)
```

来源：skillsdirectory

工作流能力诊断

📘如何使用这个技能

评论0