Use when you have a written implementation plan to execute in a separate session with review checkpoints
npx skills add algorand-devrel/algorand-agent-skills --skill "create-project"
Install specific skill from multi-skill repository
# Description
Bootstraps production-ready AlgoKit projects for Algorand dApps and smart contracts. Use when initializing new Algorand smart contract projects, setting up development environments from scratch, or scaffolding dApps with pre-configured tooling. Strong triggers include "create a new project", "initialize a new Algorand app", "start a new smart contract", "set up AlgoKit", "scaffold a dApp", "algokit init".
# SKILL.md
name: create-project
description: Bootstraps production-ready AlgoKit projects for Algorand dApps and smart contracts. Use when initializing new Algorand smart contract projects, setting up development environments from scratch, or scaffolding dApps with pre-configured tooling. Strong triggers include "create a new project", "initialize a new Algorand app", "start a new smart contract", "set up AlgoKit", "scaffold a dApp", "algokit init".
AlgoKit Project Initialization
Create new Algorand projects using AlgoKit's official templates.
Overview / Core Workflow
- Confirm project details with user (name, template, customizations)
- Run
algokit initwith appropriate flags - Handle any initialization errors
- Provide next steps for building/testing
How to proceed
- Confirm project details with user:
- Project name (directory name)
- Template choice (TypeScript or Python)
- Any customizations (
--no-git,--no-bootstrap, author name) -
For TypeScript: confirm Production preset for production projects
-
Run initialization command:
TypeScript (Production Preset):
bash
algokit init -n <project-name> -t typescript --answer preset "Production" --answer author_name "<name>" --defaults
TypeScript (Starter Preset):
bash
algokit init -n <project-name> -t typescript --answer author_name "<name>" --defaults
Python (Production Preset):
bash
algokit init -n <project-name> -t python --answer preset "Production" --answer author_name "<name>" --defaults
Python (Starter Preset):
bash
algokit init -n <project-name> -t python --answer author_name "<name>" --defaults
With custom options (no git, no bootstrap):
bash
algokit init -n <project-name> -t typescript --no-git --no-bootstrap --defaults
- Handle errors:
- Check if project directory already exists
- Verify AlgoKit is installed:
algokit --version - Ensure target directory is writable
-
Valid templates:
typescript,python,tealscript,react,fullstack,base -
Provide next steps:
cd <project-name>algokit project run buildβ Compile contractsalgokit project run testβ Run test suitealgokit localnet startβ Start local network (if deploying)algokit project run deployβ Deploy contracts to local network
Important Rules / Guidelines
- Always confirm with user before executing β Never run
algokit initwithout explicit confirmation - Default to TypeScript β Recommended for production applications
- Use Production preset β For any project because it includes testing framework and deployment scripts
- Include author name β Pass
--answer author_name "<name>"for attribution - Use
--defaultsβ Accepts all other default values for non-interactive mode
Common Variations / Edge Cases
| Scenario | Approach |
|---|---|
| Python with TypeScript deployment | --answer deployment_language "typescript" |
| Existing directory | Check and warn if directory already exists |
| No Git initialization | Use --no-git flag |
| No dependency installation | Use --no-bootstrap flag |
| Custom author name | --answer author_name "Your Name" |
| Fullstack (frontend + contracts) | Use -t fullstack template |
| React frontend only | Use -t react template |
| Standalone (no workspace) | Use --no-workspace flag |
| Initialize from example | Use algokit init example subcommand |
References / Further Reading
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.