---
title: "agent-design-tools"
description: "Skills for designing, reviewing, and improving model-facing instructions and agent workflows."
canonical: https://agentpluginsdirectory.com/plugins/agent-design-tools
last-updated: 2026-10-21
---

# agent-design-tools
Skills for designing, reviewing, and improving model-facing instructions and agent workflows.
- Slug: agent-design-tools
- Publisher: k_taka
- Repository: https://github.com/peaceroad/ai-dotfiles
- Manifest: plugins/agent-design-tools/plugin.json
- Version: 0.3.1
- Category (editorial): agent-tooling
- Skills: 3 (agent-workflow-design, prompt-design, prompt-gemini-reference)
- MCP servers: 0
- Stars: 0
- Repository created: 2026-07-17
- Repository last pushed: 2026-10-17
- Publisher type: User
- Listing: https://agentpluginsdirectory.com/plugins/agent-design-tools
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What agent-design-tools does, in the publisher's words

An Agent Plugins v1 package for designing, reviewing, and improving model-facing instructions and agent workflows. This README is a user guide; each skill's SKILL.md defines its instructions and reference-loading conditions.

Use both when a workflow needs model-facing instructions as well as execution control. For Codex skill creation or structural changes, use the built-in skill-creator for skill structure; it is provided separately from this plugin. A wording correction alone does not require workflow design, and ordinary task execution does not require these design skills merely because it uses several steps or tools.

Both skills select the model that will use the resulting instructions or workflow from the request, established conversation context, or the artifact's deployment context. For OpenAI or Codex work with no established target, they default to GPT-6 Astra.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/peaceroad/ai-dotfiles/HEAD/plugins/agent-design-tools/README.md

## Skills

- agent-workflow-design: Use when designing, reviewing, or revising agent workflows for repeated or long-running work, schedules or events, tool or agent coordination, interruption and recovery, or improvement from execution evidence. Do not use merely to execute an ordinary task, write an application loop, or edit prompt…
- prompt-design: Use when designing, reviewing, or rewriting model-facing instructions: prompts, system/developer/agent/tool instructions, prompt templates, evaluation prompts, Codex skills, Gemini prompts, Gems, and Gemini media-generation prompts. Use skill-creator first for Codex skill creation or structural cha…
- prompt-gemini-reference

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.