design-agent
crewaiinc/skills · skills.sh
Open source Repository Open in the app JSON README (API)
About
Skill publicada por crewaiinc/skills no skills.sh. Instale com: npx skills add crewaiinc/skills@design-agent
Details
- Kind
- Agent skills
- Topic
- No topic detected
- Publisher
- crewaiinc
- Origin
- skillssh
- Category
- ferramentas
- Stars
- 38
- Forks
- 17
- Open pull requests
- 2
- Last push
- 2026-06-16T16:43:58Z
- Repository state
- ativo
- Added
- 2026-08-30 15:21:25
- Updated
- 2026-09-08 15:02:18
- Origin id
crewaiinc/skills/design-agent
README
# CrewAI Skills A collection of skills for AI coding agents that teach best practices for building with [CrewAI](https://docs.crewai.com). Skills follow the [Agent Skills](https://agentskills.io/) format. ## Available Skills ### getting-started CrewAI architecture decisions and project scaffolding. Covers choosing the right abstraction (`LLM.call()` vs `Agent.kickoff()` vs `Crew.kickoff()` vs `Flow`), CLI scaffolding, YAML configuration, wiring `@CrewBase` crews, writing Flows with `@start`/`@listen`, conversational Flows with `handle_turn()`, and variable interpolation. **Use when:** - Starting a new CrewAI project - Choosing between abstraction levels - Scaffolding with `crewai create flow` - Setting up agents.yaml and tasks.yaml - Wiring crew.py or main.py - Building experimental conversational Flows - Debugging common setup issues ### design-agent CrewAI agent design and configuration. Covers the Role-Goal-Backstory framework, LLM selection, tool assignment, execution tuning (`max_iter`, `max_rpm`, `max_execution_time`), memory and knowledge sources, guardrails, and YAML vs code configuration. **Use when:** - Creating or configuring CrewAI agents - Choosing role, goal, and backstory - Assigning tools or selecting LLMs - Tuning agent parameters - Setting up knowledge sources or memory - Debugging agent behavior ### design-task CrewAI task design and configuration. Covers writing effective descriptions and expected output, task dependencies with `context`, structured output (`output_pydantic`, `output_json`, `output_file`), guardrails, human-in-the-loop review, and async execution. **Use when:** - Creating or configuring CrewAI tasks - Writing task descriptions and expected output - Setting up task dependencies - Configuring structured output formats - Adding guardrails or human review - Debugging task execution issues ## Installation In [Claude Code](https://docs.claude.com/en/docs/claude-code), add this marketplace and install the plugin: ``` /plugin marketplace add crewAIInc/skills /plugin install crewai-skills@crewai-plugins ``` The first command registers the marketplace from this repo's `.claude-plugin/marketplace.json`. The second installs the `crewai-skills` plugin from the `crewai-plugins` marketplace. To pin to a specific branch or tag: ``` /plugin marketplace add crewAIInc/skills ``` ## Skill Structure Each skill contains: - `SKILL.md` - Instructions for the agent - `references/` - Supporting documentation (tools catalog, MCP servers, structured output patterns, etc.) ## License MIT