# Runnable

> Use when the user asks for a one-word acknowledgement drill — replies with exactly the word the drill names.

- Skill: `ai-creed/runnable` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add ai-creed/runnable`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ai-creed/runnable/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Docs & Writing
- Author: ai-creed (https://skillmd.com/u/ai-creed)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/ai-creed/runnable

---


# runnable

## Intent

Give the eval-run exercise a minimal, valid target: a skill whose eval suite
is deterministic-clean and cheap to execute, with one case that fails its
expectation by design so a runner has a real failure to report.

## Inputs

- A drill prompt naming what to say.

## Preconditions

- None.

## Procedure

1. Read the drill prompt.
2. Reply with exactly what it asks for, nothing more.

## Output

- The requested reply, verbatim.

## Examples

Input: `Say the word apple.`
Output: `apple`

## Anti-patterns

- Adding commentary around the requested reply.