AWS Skills
AI agent skills for deploying and managing infrastructure on Amazon Web Services.
Skills
codepipeline-github-status-setup
Sets up GitHub commit status checks for AWS CodePipeline → Elastic Beanstalk deployments.
| Component | Detail |
|---|---|
| Trigger | EventBridge rule on CodePipeline execution state changes |
| Handler | Python 3.12 Lambda function |
| Secret storage | SSM Parameter Store (SecureString) |
| Infrastructure | CloudFormation stack (IAM role, Lambda, EventBridge rule, Log Group) |
| GitHub integration | Statuses API — aws/elastic-beanstalk/<env> context per pipeline |
Covers the full lifecycle:
- Phase 1 — Collect inputs (GitHub token, repo, AWS profile, region, resource prefix, pipeline mappings)
- Phase 2 — Generate infrastructure files (
lambda_function.py,template.yml,deploy.sh) - Phase 3 — Store GitHub PAT in SSM Parameter Store
- Phase 4 — Deploy CloudFormation stack and upload Lambda code
- Phase 5 — Verify EventBridge → Lambda wiring
Each covered commit shows one status check per environment — pending while deploying, success or failure on completion — linking directly to the CodePipeline execution timeline in the AWS console.
Installation
Via skills CLI (recommended)
npx skills add UEM-Group-Websites/aws-skills
From a release (manual)
Download the .skill file for a specific skill from the Releases page, then install it:
# Example for codepipeline-github-status-setup
unzip codepipeline-github-status-setup.skill -d ~/.agents/skills/
ln -s "../../.agents/skills/codepipeline-github-status-setup" \
~/.claude/skills/codepipeline-github-status-setup
Claude Code (manual)
# 1. Clone this repo
git clone https://github.com/UEM-Group-Websites/aws-skills.git /tmp/aws-skills
# 2. Copy the skill(s) you want
mkdir -p ~/.agents/skills/codepipeline-github-status-setup
cp -r /tmp/aws-skills/codepipeline-github-status-setup/. \
~/.agents/skills/codepipeline-github-status-setup/
# 3. Create the symlink Claude Code needs
ln -s "../../.agents/skills/codepipeline-github-status-setup" \
~/.claude/skills/codepipeline-github-status-setup
Restart Claude Code — the skill will appear automatically.
Usage
Once installed, trigger the skill by telling Claude Code something like:
- “Set up GitHub commit status checks for my CodePipeline deployments”
- “Add deployment status badges to our GitHub commits from CodePipeline”
- “I want to see pending/success/failure on commits when Elastic Beanstalk deploys”
- “Wire up CodePipeline to the GitHub Statuses API”
Claude will walk you through each phase, collecting your project’s real values for every placeholder before generating and deploying anything.
Repository layout
aws-skills/
└── codepipeline-github-status-setup/
├── SKILL.md # Skill instructions and input collection workflow
└── templates/
├── lambda_function.py # Python Lambda template
├── template.yml # CloudFormation template
└── deploy.sh # Package and deploy script
Releases
Tagged releases (e.g. v1.0.0) automatically package every skill as a .skill file and attach it to the GitHub Release. A .skill file is a standard zip archive — you can inspect or extract it with any zip tool.
License
MIT