AWS Skill

A comprehensive AWS integration skill for Skill Engine, providing access to S3, EC2, and Lambda services.

Quick Start

# Run directly from directory (no build needed!)
skill run ./examples/aws-skill s3-list bucket=my-bucket

# List running EC2 instances
skill run ./examples/aws-skill ec2-list state=running

# Invoke a Lambda function
skill run ./examples/aws-skill lambda-invoke function=my-function payload='{"key":"value"}'

Features

• Zero Configuration: Just write JavaScript and run
• Auto-Compilation: Runtime compiles to WASM on first use (~3 seconds)
• Cached Execution: Subsequent runs use cached WASM (<100ms startup)
• Secure Credentials: AWS keys stored in system keychain
• Multi-Account: Support for multiple AWS accounts via instances

Available Tools

S3 Operations

• s3-list - List objects in a bucket
• s3-upload - Upload files to S3
• s3-download - Download files from S3

EC2 Operations

• ec2-list - List and filter EC2 instances

Lambda Operations

• lambda-invoke - Invoke Lambda functions synchronously or asynchronously

Configuration

Method 1: Config File

Create skill.config.toml:

[config]
aws_access_key_id = "AKIAIOSFODNN7EXAMPLE"
aws_secret_access_key = "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY"
region = "us-east-1"

Method 2: Interactive Wizard

# Install skill first
skill install ./examples/aws-skill --instance prod

# Configure interactively
skill config aws-skill --instance prod
# Follow prompts to enter credentials

Method 3: Environment Variables

export SKILL_AWS_ACCESS_KEY_ID="AKIA..."
export SKILL_AWS_SECRET_ACCESS_KEY="..."
export SKILL_REGION="us-east-1"

skill run ./examples/aws-skill s3-list bucket=my-bucket

Multi-Account Usage

Support for multiple AWS accounts:

# Install for production account
skill install ./examples/aws-skill --instance prod
skill config aws-skill --instance prod
# Enter production credentials

# Install for staging account
skill install ./examples/aws-skill --instance staging
skill config aws-skill --instance staging
# Enter staging credentials

# Use specific account
skill run aws-skill --instance prod s3-list bucket=prod-data
skill run aws-skill --instance staging s3-list bucket=staging-data

Examples

S3 Examples

# List all objects in a bucket
skill run ./examples/aws-skill s3-list bucket=my-bucket

# List objects with prefix (like a folder)
skill run ./examples/aws-skill s3-list bucket=my-bucket prefix=logs/2024/

# Upload a file
skill run ./examples/aws-skill s3-upload \
  bucket=my-bucket \
  key=uploads/file.txt \
  file=./local-file.txt

# Download a file
skill run ./examples/aws-skill s3-download \
  bucket=my-bucket \
  key=data/report.pdf \
  output=./report.pdf

EC2 Examples

# List all instances
skill run ./examples/aws-skill ec2-list

# List only running instances
skill run ./examples/aws-skill ec2-list state=running

# Filter by tag
skill run ./examples/aws-skill ec2-list tag=Environment=production

Lambda Examples

# Invoke function with no payload
skill run ./examples/aws-skill lambda-invoke function=my-function

# Invoke with JSON payload
skill run ./examples/aws-skill lambda-invoke \
  function=data-processor \
  payload='{"action":"process","items":["a","b","c"]}'

# Invoke asynchronously
skill run ./examples/aws-skill lambda-invoke \
  function=email-sender \
  payload='{"to":"user@example.com","subject":"Hello"}' \
  async=true

Security

Credential Storage

• Credentials are stored in your system's secure keychain:

  • macOS: Keychain Access
  • Windows: Credential Manager
  • Linux: Secret Service API (gnome-keyring, KWallet)

• Secrets are encrypted at rest

• Never logged or printed to console

• Cleared from memory after use

IAM Best Practices

1. Use IAM Users, not Root: Create dedicated IAM users for CLI access
2. Least Privilege: Grant only the permissions you need
3. Rotate Keys: Change access keys regularly
4. Enable MFA: Add multi-factor authentication to your IAM user

Required IAM Permissions

See SKILL.md for detailed IAM policy examples.

Development

This skill is written in pure JavaScript and can be modified directly:

# Edit the skill
vim examples/aws-skill/skill.js

# Run immediately - automatically recompiles if changed
skill run ./examples/aws-skill s3-list bucket=test

Adding New AWS Services

1. Import the AWS SDK client (if using real AWS SDK)
2. Add tool definition to getTools()
3. Implement handler function
4. Add to switch statement in executeTool()
5. Update SKILL.md documentation

Real AWS SDK Implementation

This example uses simulated responses for demonstration. To connect to real AWS:

1. Uncomment the AWS SDK imports at the top of skill.js
2. Ensure AWS SDK packages are available during compilation
3. Replace simulated responses with real SDK calls

The skill structure is already set up for real AWS integration.

Troubleshooting

"Credentials not configured"

• Run skill config aws-skill to set up credentials
• Or create a skill.config.toml file with your credentials

"Region not specified"

• Set region in your config file
• Or use SKILL_REGION environment variable

"Access Denied" errors

• Check IAM permissions for your credentials
• Verify the resource (bucket, instance, function) exists
• Ensure you're using the correct region

Documentation

See SKILL.md for comprehensive documentation including:

• What is AWS and when to use this skill
• Detailed tool reference
• Security best practices
• IAM permission requirements
• Troubleshooting guide

License

MIT License - Part of Skill Engine project