agent-design-coach

agent-design-coach is a skill for Claude Code, Codex from dadwadw233/claude-code-philosophy. It costs 89 tokens per session (1,181 once invoked), scanned A, original, MIT.

A design guide for building useful and controllable software agents. It helps define the user’s goal, tools, permissions, memory, recovery behavior, and evaluation approach.

In plain words
What is it for?
Use it when designing an agent architecture, tool system, memory strategy, permission model, interaction loop, recovery plan, or evaluation process.
Why use it?
It makes an agent’s behavior and limits explicit, which helps people understand, debug, and safely improve it.

Skill for Claude CodeCodex

Written for Claude Code and Codex: shipped in a Claude Code plugin, but also agents/openai.yaml present.

Part of the claude-code-philosophy plugin — 1 skill shipped together

Good fit Use it when designing an agent architecture, tool system, memory strategy, permission model, interaction loop, recovery plan, or evaluation process.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/dadwadw233/claude-code-philosophy/agent-design-coach
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add dadwadw233/claude-code-philosophy --skill agent-design-coach
Clone the repo
git clone --depth 1 https://github.com/dadwadw233/claude-code-philosophy

Made for: Claude Code, Codex.

Or install claude-code-philosophy, the plugin that ships this one along with the rest of its 1 skill.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for agent-design-coach

README.md
[![agentmods](https://agentmods.dev/badge/skills/dadwadw233/claude-code-philosophy/agent-design-coach/github.svg)](https://agentmods.dev/skills/dadwadw233/claude-code-philosophy/agent-design-coach)
Your own site
<a href="https://agentmods.dev/skills/dadwadw233/claude-code-philosophy/agent-design-coach"><img src="https://agentmods.dev/badge/skills/dadwadw233/claude-code-philosophy/agent-design-coach/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for agent-design-coach

Your own site · 80×15
<a href="https://agentmods.dev/skills/dadwadw233/claude-code-philosophy/agent-design-coach"><img src="https://agentmods.dev/badge/skills/dadwadw233/claude-code-philosophy/agent-design-coach.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 89 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,181 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00089 $0.01181
Opus 5 $0.00044 $0.00590
Sonnet 5 $0.00018 $0.00236
Haiku 4.5 $0.00009 $0.00118

Measured 9d ago against content hash b72fda29f65b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

agent-design-coach scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/agent-design-coach/SKILL.md · 192 lines

How it starts

The opening of the file, as written. The whole thing — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Design Coach

Design agents as products, not as prompt demos.

This skill helps you turn fuzzy agent ideas into systems that are:

  • useful to a real user
  • controllable by humans
  • explicit about tools and permissions
  • resilient across long-running work
  • understandable enough to debug and improve

What to optimize for

Prefer systems that are:

  • legible over magical
  • composable over monolithic
  • recoverable over one-shot
  • bounded over all-powerful
  • product-useful over benchmark-impressive

Quick start

  1. Identify the user, the job, and the failure cost.
  2. Decide whether the system needs a real harness or only a workflow/script.
  3. Define the execution loop before adding more features.
  4. Treat tools, memory, permissions, and recovery as first-class design surfaces.
  5. End with a concrete design, risks, and an evaluation plan.

Output contract

When this skill is active, prefer delivering outputs with these sections:

  1. Goal and user value
  2. Recommended system shape
  3. Harness and execution loop
  4. Tools and capability boundaries
  5. Context and memory strategy
  6. Safety, permissions, and recovery
  7. UX and human control
  8. Failure modes and evaluation
  9. Immediate next implementation steps

If the user asks for a review, prioritize flaws, weak assumptions, and missing controls before praise.

Workflow

1. Start from the user job, not the model

Clarify:

  • who the user is
  • what task they are actually delegating
  • what “good” looks like
  • what damage happens when the agent is wrong, slow, or overconfident

If the task is not worth repeated delegation, challenge whether an agent is needed at all.

2. Decide the minimum viable harness

Do not jump from “call a model” to “full autonomous agent.”

Pick the smallest useful runtime shape:

  • one-shot assistant
  • guided multi-step copilot
  • tool-using loop
  • long-running agent with persistence and recovery
  • multi-agent system

Only escalate complexity when the product actually needs it.

Read the full file on GitHub · 192 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 192 lines · 89 tokens per session scan A b72fda29f65b

Subscribe to this mod's changes

agent-design-coach is a skill published in the GitHub repository dadwadw233/claude-code-philosophy (5 stars, last pushed 5mo ago), licensed MIT. It adds 89 tokens to every session and 1,181 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

rubber-duck

Adversarial "rubber duck" review that turns explaining-out-loud into a hallucination check. The main session is the PRESENTER (it did the work — a design doc, investigation, or analysis — and holds the real reasoning) and reconstructs the topic to a LISTENER — a spawned subagent pinned to a DIFFERENT-vendor model that…

kirodotdev/KiroCrew · 161 tokens

ai-ml-development

AI and machine learning development with PyTorch, TensorFlow, and LLM integration. Use when building ML models, training pipelines, fine-tuning LLMs, or implementing AI features.

travisjneuman/.claude · 43 tokens

authentication-patterns

OAuth 2.0, JWT, SSO, MFA, NextAuth/Clerk/Supabase Auth implementation patterns.

travisjneuman/.claude · 28 tokens

case-interview-practice

Interactive consulting case interview practice with structured frameworks, feedback mechanisms, and progressive difficulty. Use when preparing for management consulting interviews, case competitions, or business problem-solving exercises.

travisjneuman/.claude · 39 tokens

finance

Financial analysis expertise for financial modeling (DCF, LBO, M&A), valuation, financial statement analysis, capital allocation, treasury management, and corporate finance decisions. Use when building financial models, analyzing statements, or making investment decisions.

travisjneuman/.claude · 48 tokens

i18n-localization

Internationalization and localization for global applications. Use when adding multi-language support, handling regional formats, or preparing apps for global markets.

travisjneuman/.claude · 32 tokens