SkillFoxx

Agentic Code skill

AGENTS.md instructions and skills with a test-first process that scales to the task, for Cursor, Codex CLI, Gemini CLI.

  • 51 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • The instructions let the agent run tests and edit repository files under the agreed process.
All reasons and checks

shinpr/agentic-codeAuthor:Shinsuke Kagawa

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/agentic-code

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The instructions let the agent run tests and edit repository files under the agreed process.

This is third-party code. Review the repository files before installing.

What it does

Agentic Code adds an `AGENTS.md` file and an `.agents/` directory to a repository. You describe a change in plain words and the agent follows a process sized to the task. A small fix goes straight to implementation, while a change spanning several parts of the system gets a design and a plan. The rules target typical agent failures. Scope stays within the request: the agent records what you want and what you ruled out. Questions are limited to requirements, permissions and irreversible actions. Before finishing, the agent runs the repository tests and proves each behavior change with one test. Review feedback is weighed before it becomes work. `.agents/skills/` holds guidance on coding, testing and documentation, and `npx agentic-code skills` installs only those.

Who it is for. Teams working with Cursor, Codex CLI or Gemini CLI who want one process with tests and bounded scope.

Good fit when

  • A new project should start with a process, test discipline and agent rules.
  • An existing repository needs `AGENTS.md` and an `.agents` directory.
  • The agent often builds extras or claims completion without proof.
  • You want only the coding and testing guidance in the Cursor or Codex skills directory.

Not a fit when

  • A well-scoped fix, throwaway experiment or one-off script.
  • You need an application generator.
  • You need a Codex-native setup with custom agents and isolated review contexts.

Example request

Add an API endpoint for user search without changing the existing response contract, following the AGENTS.md process.

Limitations

The files add process and context to every session, so for small tasks a plain agent session is cheaper. Node.js 22 or newer is required. How each tool discovers instructions and skills depends on its version. The guidance is language-agnostic with extra references for TypeScript, and in other languages the agent relies on the commands and conventions of the repository itself. The files set the order in which the agent works and create no application.

How to disable. Delete `AGENTS.md` and the `.agents` directory from the repository, and the skills from the tool directory if you installed them with the `skills` command.

FAQ

How do I start a new project?

Run `npx agentic-code my-project`. It creates a directory with `AGENTS.md` and `.agents`, initializes Git and makes a first commit.

How do I add it to an existing repository?

Copy `AGENTS.md` and the `.agents` directory into the root and check build and test settings. Your repository commands take precedence.

How do I install only the skills?

Run `npx agentic-code skills --codex` or `--cursor`, add `--project` for the current project or `--path` for a custom directory.

Which tools read the files?

Tools that read a repository-level `AGENTS.md`, with Cursor, Codex CLI and Gemini CLI as common examples.

When should I skip it?

For well-scoped fixes and one-off scripts, where the process adds context and a plain session costs less.

Discussion

No questions or reviews yet.

Skills

Agentic OS adds a five-step workflow and checks in git hooks and CI so that a step without evidence does not count as done

WorkflowExtended access234

A library of 30-plus instruction skills for coding agents: planning, building, testing, hallucination guards and shipping

SkillMinimal access217

The author's personal set: shared agent rules and dozens of skills for GitHub, the browser, macOS, messaging and releases, plus a sync script

SkillBroad access7.5K

A skill that creates and updates AGENTS.md files per the agents.md spec: a thin root file plus separate files for each subsystem

PluginMinimal access89

Two skills for coding agents: wtf audits a project before shipping, better-codex keeps the agent disciplined while working on code

SkillExtended access1.4K

Builder Skills

Skills for agents

A set of small skills from Builder.io for agent work habits: visual plans and diff recaps, docs-first checks, cross-agent review and clean status updates

SkillExtended access4.7K

README in short

The instructions define a process that grows with the change and target extra work, unneeded questions and premature completion claims. You can install everything or only the skills.

README badge

Original SKILL.md

---
name: ai-development-guide
description: "Applies root-cause analysis, impact inspection, and completeness checks without expanding the requested scope. Use when fixing bugs, refactoring, reviewing quality, or resolving unexpected implementation evidence."
---

# AI Development Guide

## Core Rule

Correct the invalid source before adding compensating layers. A workaround, fallback, abstraction, or new artifact is justified only when current evidence shows that removing or fixing the cause cannot satisfy the approved outcome.

## Investigation

Inspect the smallest relevant path that can establish:

- expected versus actual behavior;
- the first failing boundary;
- callers, contracts, and data or control flow affected;
- existing equivalent behavior and representative repository patterns;
- tests and checks that can reproduce or observe the issue;
- unknowns that could change the fix.

Mark findings as observed, inferred, or unknown. Expand inspection when new evidence points to another affected boundary, not to satisfy a fixed coverage checklist.

## Root-Cause Resolution

1. Reproduce or otherwise observe the failure.
2. Identify the earliest invalid assumption, state, or contract.
3. Determine whether existing code, configuration, or documentation already provides the correct mechanism.
4. Fix or remove the invalid source.
5. Add a focused regression proof when it protects the demonstrated failure from recurrence.

Use a minimal reproduction or iterative why-analysis when it changes the cause decision. Do not create either as a ceremonial artifact.

## Existing Mechanism Decision

When similar functionality exists:

- reuse it when its responsibility, lifecycle, and contract fit the current outcome;
- extend it when the extension preserves its responsibility and remains cheaper than a parallel mechanism;
- replace or remove it when evidence shows it is broken or unused;
- create a new mechanism only when the current ones cannot satisfy the requirement.

Technical debt discovered outside the current outcome remains a finding, not an automatic ADR or implementation task.

Report a problem