skills.rest
catlog22
catlog22COMMUNITY
Explore all skills by this author →
GitHub

ccw-cli-tools

Orchestrates multi-CLI AI tool execution with unified prompt templates and tag-based routing.

2.1k|165|Updated Sep 7, 2025
Search Tags:#multi-model#CODEX#GEMINI#prompt template#cli orchestration#Code review#Code Analysis
One-click install
npx skills add https://github.com/catlog22/Claude-Code-Workflow --skill ccw-cli-tools
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ccw-cli-tools
Source: https://github.com/catlog22/Claude-Code-Workflow/tree/main/.codex/skills/ccw-cli-tools
Command: npx skills add https://github.com/catlog22/Claude-Code-Workflow --skill ccw-cli-tools

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinating multiple AI CLI tools (Gemini, Codex, Qwen, Claude, OpenCode) for code analysis and implementation tasks requires consistent prompt structures, correct tool selection, and safe permission handling, which is error-prone when done manually.

Core Features & Use Cases

  • Configuration-Driven Tool Selection: Reads ~/.claude/cli-tools.json to route tasks to enabled tools by capability tags, with primary/secondary model fallback chains.
  • Unified 6-Field Prompt Template: Standardizes PURPOSE, TASK, MODE, CONTEXT, EXPECTED, and CONSTRAINTS fields for every CLI invocation.
  • Permission-Gated Modes: Enforces analysis (read-only), write (file modification), and review (git-aware, Codex-only) modes with explicit authorization.
  • Use Case: Run a security audit by invoking ccw cli with the analysis-assess-security-risks rule template against the auth module, then resume the session in write mode to implement the recommended fixes.

Quick Start

Ask the AI to analyze the authentication module for security vulnerabilities using the ccw cli with the security assessment rule template in analysis mode.

Frequently Asked Questions about ccw-cli-tools

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a code analysis with the ccw cli command?▼

Use `ccw cli -p "<PROMPT>" --tool <tool> --mode analysis --rule <template>` with the six-field prompt structure covering PURPOSE, TASK, MODE, CONTEXT, EXPECTED, and CONSTRAINTS. Analysis mode is read-only and safe for auto-execution.

How does the skill choose between Gemini, Codex, Qwen, and Claude?▼

Tool selection is configuration-driven from `~/.claude/cli-tools.json`. An explicit `--tool` flag wins; otherwise the task's required capability tags are matched against each enabled tool's tags, falling back to the first enabled tool.

What is the difference between analysis, write, and review modes?▼

Analysis mode is read-only and safe for auto-execution. Write mode permits creating, modifying, and deleting files and requires explicit `--mode write`. Review mode performs git-aware review of uncommitted changes, branches, or commits and is supported only by Codex.

Can I resume a previous CLI session for multi-round tasks?▼

Yes, use `--resume` to continue the last session or `--resume <id>` for a specific session. You can also merge multiple sessions with comma-separated IDs, which supports multi-round planning and multi-model collaboration.

What happens when the primary model or tool fails?▼

The fallback chain first tries the tool's secondaryModel from the configuration, then the next enabled tool with matching tags, and finally the first enabled tool in the configuration as the default.

When should I not use write mode with ccw cli?▼

Avoid write mode unless the user explicitly requests file creation or modification, since authorization is single-use and does not carry over between executions. Default to analysis mode for exploration, review, and diagnosis tasks.

Related Skills

Explore More Software Engineering Skills→
vuminhieuvuminhieu

mcp-management

Discover, analyze, and execute MCP capabilities across configured servers.

Community
Advanced
ArcReelArcReel
4.3k

implement

Implement work from a spec or tickets using TDD and code review.

Official
Basic
neo-picasso-2112neo-picasso-2112

code-review-excellence

Establish structured code-review standards with checklists and templates for pull requests.

Community
Advanced
gqf2008gqf2008

blackbox

Orchestrate multiple AI models for coding tasks via the Blackbox CLI.

Community
Advanced