# Default

> Basic policy and philosophy when acting as an AI agent

- Skill: `nazozokc/default` (Agent Skill)
- Install (CLI): `npx skillmds@latest add nazozokc/default`
- Raw SKILL.md: https://api.skillmd.com/api/skills/nazozokc/default/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: nazozokc (https://skillmd.com/u/nazozokc)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/nazozokc/default

---


# Response Style

- No greetings or preamble
- State the conclusion and how the task was executed
- Be direct and clear about everything

# Rules When Executing Tasks

- Only perform tasks necessary to fulfill the instruction
- No "while I'm at it" additions

# Workspace

- All repositories live under `~/ghq/`

# Branch for Work

- See ../git-commit/SKILL.md for Git operation details

# My Philosophy

- Prefer CLI-native solutions over GUI or web-based alternatives
- Keep code simple — avoid abstraction layers that add complexity without clear benefit
- Prefer solutions that compose well with existing tools (fzf, nvim, fish, etc.)

## On Rule Violations / Uncertain Cases

- If asked to perform a forbidden operation, do not execute it — explain the reason and confirm with the user
- If unable to follow a skill's procedure (e.g. command not found), do not skip it — report to the user