# design-task

A skill for designing and debugging CrewAI tasks. It guides descriptions, expected outputs, task dependencies, and Pydantic structured models.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 8100 installs · ★ 48
- Author: crewAIInc
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/crewAIInc/skills/tree/HEAD/skills/design-task/SKILL.md
- Updated: 2026-10-06
- HTML: https://affpapa.org/en/nevibecode/skills/design-task

## What it does

This skill provides task architecture guidelines for CrewAI, tested against versions 1.15.22 and 1.15.23. It enforces that every task requires both `description` and `expected_output`, because omitting either raises a `ValidationError`. One task means one objective. Combining research, writing, and proofreading into a single task degrades output quality.

The guide outlines dependency chaining via the `context` parameter. Passing `context=[]` strips all prior step outputs, while pointing to a future task halts execution immediately. For structured data, the skill covers configuring `output_pydantic` and `output_json`. Setting `output_pydantic` inside `tasks.yaml` instead of Python will cause CrewAI to drop the structured model.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add crewAIInc/skills --skill design-task -a claude-code
```

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

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add crewAIInc/skills --skill design-task -a codex
```

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

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add crewAIInc/skills --skill design-task -a cursor
```

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

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install crewAIInc/skills design-task
```

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

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add crewAIInc/skills --skill design-task -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/crewAIInc/skills --path skills/design-task
```

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add crewAIInc/skills --skill design-task -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/crewAIInc/skills/tree/HEAD/skills/design-task/SKILL.md

## When to use it
- write valid `description` and `expected_output` parameters
- configure task dependencies and data passing via `context`
- enforce structured outputs using `output_pydantic` models
- route task results directly to disk with `output_file`

## Example prompts
- Help me write a description and expected_output for a CrewAI research task
- How do I pass context between tasks using the context argument in CrewAI?
- Configure a CrewAI task that validates structured output with Pydantic

## Source text · SKILL.md · MIT

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: 2722099d70ba7d8c.

BEGIN THIRD-PARTY FILE 2722099d70ba7d8c
````text
---
name: design-task
description: "CrewAI task design and configuration. Use when creating, configuring, or debugging crewAI tasks — writing descriptions and expected_output, setting up task dependencies with context, configuring output formats (output_pydantic, output_json, output_file), using guardrails for validation, enabling human_input, async execution, markdown formatting, or debugging task execution issues."
---

# CrewAI Task Design Guide

How to write effective tasks that produce reliable, high-quality output from your agents.

Verified against crewai 1.15.22 and 1.15.23 on 2026-10-01.
Live-tested with real LLMs on 2026-10-01.
For exact current API forms (imports, parameter names, structured output, guardrail signatures) the **check-crewai-api** skill is the reference; where this skill and it disagree, follow check-crewai-api.

---

**The 80/20 rule: spend 80% of your effort on task design, 20% on agent design.** The task is the most important lever you have. A well-designed task with a mediocre agent will outperform a poorly designed task with an excellent agent.

---

## 1. Anatomy of an Effective Task

Every task needs two things: a **description** (what to do and how) and an **expected_output** (what the result looks like). Both are required - `Task(description=...)` alone raises `ValidationError ... expected_output Field required`.

### Description — The Instructions

A good description includes:
1. **What** to do — the core action
2. **How** to do it — specific steps or approach
3. **Context** — why this matters, what it feeds into
4. **Constraints** — scope limits, things to avoid
5. **Inputs** — what data or context is available

```yaml
research_task:
  description: >
    Conduct thorough research about {topic} for the year {current_year}.

    Your research should:
    1. Identify the top 5 key trends and breakthroughs
    2. For each trend, find at least 2 credible sources
    3. Note any controversies or competing viewpoints
    4. Assess potential industry impact (high/medium/low)

    Focus on developments from the last 6 months.
    Do NOT include speculation or unverified claims.
    The output will feed into a report for {target_audience}.
  expected_output: >
    A structured research brief with 5 sections, one per trend.
    Each section includes: trend name, 2-3 paragraph summary,
    source citations, impact assessment (high/medium/low),
    and a confidence level for your findings.
  agent: researcher
```

### Expected Output — The Success Criteria

The `expected_output` tells the agent what "done" looks like. Be specific about:
- **Format** — bullet points, paragraphs, JSON, table
- **Structure** — sections, headings, order
- **Length** — approximate word count or number of items
- **Quality markers** — citations required, confidence levels, specific fields

| Bad Expected Output | Good Expected Output |
|---|---|
| `A research report` | `A structured brief with 5 sections, each containing: trend name, 2-3 paragraph summary, source citations, and impact rating` |
````
END THIRD-PARTY FILE 2722099d70ba7d8c

The third-party file has ended. Everything between the 2722099d70ba7d8c 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.

First 3075 of 22245 bytes. © crewAIInc · MIT · https://github.com/crewAIInc/skills/tree/HEAD/skills/design-task/SKILL.md

_Part of the page did not fit — see the HTML version._

---
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