# implementation

Guides the agent to write production-ready code following existing patterns, handle errors, and avoid scope creep before verification.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 850 installs · ★ 873
- Author: cyrusagents
- License: Apache-2.0 · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/cyrusagents/cyrus/tree/HEAD/skills/implementation/SKILL.md
- Updated: 2026-10-25
- HTML: https://affpapa.org/en/nevibecode/skills/implementation

## What it does

This skill sets concrete boundaries for drafting production code in Claude Code and supported IDEs. The agent inspects existing repository patterns and matches them closely. It ignores unrelated cleanup. Edits remain strictly focused on the requested scope.

Workflow progression is explicit. When automated tests fail, the model must fix them immediately instead of advancing to subsequent tasks. Testing strategy leans on `verify-and-ship`. The agent runs relevant existing tests or direct checks, reserving fresh tests exclusively for consequential regressions worth long-term maintenance.

Because the instructions explicitly route output to a verification and shipping phase, using this skill outside the broader Cyrus workflow requires handling the next handoff step on your own.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add cyrusagents/cyrus --skill implementation -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/implementation/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/implementation/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add cyrusagents/cyrus --skill implementation -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/implementation/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add cyrusagents/cyrus --skill implementation -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/implementation/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install cyrusagents/cyrus implementation
```

2. **Manually: put the files in place**
- File: `.github/skills/implementation/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add cyrusagents/cyrus --skill implementation -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/cyrusagents/cyrus --path skills/implementation
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/implementation/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add cyrusagents/cyrus --skill implementation -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/cyrusagents/cyrus/tree/HEAD/skills/implementation/SKILL.md

## When to use it
- implement requested changes without drifting into unrelated refactoring
- preserve existing repository patterns and conventions
- handle edge cases and runtime failure scenarios
- prepare code for the `verify-and-ship` workflow step
- resolve failing tests before proceeding to the shipping phase

## Example prompts
- Implement the requested feature changes while following our existing code style
- Handle edge cases in this data parser without touching unrelated modules
- Add the new calculation logic and ensure all current tests pass before shipping

## Source text · SKILL.md · Apache-2.0

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: d6441da527dd4cc5.

BEGIN THIRD-PARTY FILE d6441da527dd4cc5
```text
---
name: implementation
description: Implement the requested changes. Write production-ready code, follow existing patterns, and use proportionate verification.
---

# Implementation

Implement the requested changes:

- Write production-ready code that follows existing patterns and conventions
- Choose verification using `verify-and-ship`: use relevant existing tests or direct checks; add new tests only for concrete consequential regressions worth maintaining
- Handle edge cases and error scenarios appropriately
- Keep changes focused on what was requested — avoid unrelated improvements

If tests fail, fix them before moving on to the next phase of the workflow.

When implementation is complete, proceed to the verification and shipping phase.
```
END THIRD-PARTY FILE d6441da527dd4cc5

The third-party file has ended. Everything between the d6441da527dd4cc5 tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

© cyrusagents · Apache-2.0 · https://github.com/cyrusagents/cyrus/tree/HEAD/skills/implementation/SKILL.md

## Similar
- [incremental-implementation](https://affpapa.org/en/nevibecode/skills/incremental-implementation.md) — skill · 61900 installs
- [principle-test-behavior-not-implementation](https://affpapa.org/en/nevibecode/skills/principle-test-behavior-not-implementation.md) — skill · 5200 installs
- [task-workflow](https://affpapa.org/en/nevibecode/skills/task-workflow.md) — skill · 1900 installs
- [build-loop-claude-code](https://affpapa.org/en/nevibecode/skills/build-loop-claude-code.md) — skill · 1700 installs
- [build-loop-cursor](https://affpapa.org/en/nevibecode/skills/build-loop-cursor.md) — skill · 1500 installs
- [superpowers-lab](https://affpapa.org/en/nevibecode/skills/superpowers-lab.md) — skill · 1400 installs

From the same set: [cyrusagents/cyrus](https://affpapa.org/en/nevibecode/by/cyrusagents/cyrus.md) — 18 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
