login
terry-li-hm

claude-code

by @terry-li-hm in AI & LLM
0
0
# Install this skill:
npx skills add terry-li-hm/skills --skill "claude-code"

Install specific skill from multi-skill repository

# Description

High-reasoning agentic operations via the Claude Code CLI (Opus 4.5). MUST use --dangerously-skip-permissions for all background tasks. Use for complex analysis and strategic decisions.

# SKILL.md

name: claude-code
description: High-reasoning agentic operations via the Claude Code CLI (Opus 4.5). MUST use --dangerously-skip-permissions for all background tasks. Use for complex analysis and strategic decisions.
user_invocable: true

Claude Code

Programmatic bridge to the Claude Code CLI, delegating high-stakes reasoning tasks to Opus 4.5.

When to Use

  • Strategic Analysis: Evaluating job fit, synthesizing interview signals, or career planning.
  • Structural Coding: Large-scale refactors or complex debugging that exceeds GLM-4.7 capabilities.
  • Max Reasoning: Any task where the "Consultant" persona is required over the "Mechanical" one.

Usage Patterns

Task Type Command Pattern Reference Skill
Job Eval claude --dangerously-skip-permissions "evaluate-job <URL> --vault ~/notes" evaluate-job
Strategic Review claude --dangerously-skip-permissions "Analyze [[Active Pipeline]] for blockers" vault-pathfinding
Code Audit claude --dangerously-skip-permissions "Run a security audit on ./src" codebase-cleanup

Best Practices

  • Zero-Interaction: Always include --dangerously-skip-permissions in exec calls to prevent hanging on TTY prompts during background/cron runs.
  • Explicit Model: Instruct Claude Code within the prompt string to use "Opus" or "4.5" for high-stakes reasoning.
  • Reporting: Always instruct Claude to write output to a specific vault path so it can be retrieved by OpenClaw.
  • llm-routing β€” Guidance on when to escalate to Claude Code vs OpenCode.
  • evaluate-job β€” Primary consumer of the Claude Code bridge.
  • vault-pathfinding β€” Standard paths for reading/writing results.

# Related Skills

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

⚑ Amp πŸš€ Antigravity πŸ€– Claude Code πŸ¦€ Clawdbot πŸ“ Codex ▢️ Cursor πŸ€– Droid πŸ’Ž Gemini CLI πŸ™ GitHub Copilot πŸͺΏ Goose πŸ“Š Kilo Code πŸ”§ Kiro CLI πŸ’» OpenCode 🦘 Roo Code 🌲 Trae πŸ„ Windsurf

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.