AWSBestPracticesSkill
title: AWS Best Practices Skill type: tool created: 2026-07-11 last_updated: 2026-07-11 related: ["radar/platforms/AnthropicSkillsRepository", "radar/techniques/AWSSaaSLens"] sources: ["https://github.com/ferdinandobons/AWSBestPracticesSkill"] radar_quadrant: Tools radar_ring: Assess radar_position: inner
An unofficial, community-maintained skill for Claude Code and OpenAI Codex that compiles AWS best practices for 208 services across 217 files, organized by the six Well-Architected Framework pillars: security, reliability, performance efficiency, cost optimization, operational excellence, and sustainability.
An AI coding agent asked to secure an S3 bucket or size an RDS instance has two default options: answer from training data, which goes stale as AWS services evolve, or run a live web search for every question, which costs time and tokens and produces answers that vary in quality. This skill replaces both with a pre-built, locally-cached catalog the agent can consult directly.
Each AWS service and cross-service topic gets its own markdown file, with every recommendation linked back to official AWS documentation rather than paraphrased from memory. Files carry context tags indicating when a given practice applies, and the skill responds to natural-language prompts such as "secure my S3 bucket" by matching against the relevant service files. An automated maintenance workflow tracks newly released AWS services and refreshes entries so the catalog does not drift out of date.
The repository is young, created 2026-06-29, but shows unusually disciplined engineering for a community project: MIT license, 34 stars, zero open issues, versioned releases (v0.1.5 and beyond), validation scripts that check structural integrity and link health, and transparent tracking of generation cost (reported at roughly 30.51M tokens across 481 workflow runs). Confirmed via GitHub: last pushed 2026-07-06.
AWS Best Practices Skill sits in the Assess ring of the Tools quadrant, at inner position. First noted via its GitHub repository (bookmarked 2026-07-11). No first-person production use yet, so Assess rather than Trial. Inner position reflects direct complementarity with radar/platforms/AnthropicSkillsRepository, the canonical Claude Code skill ecosystem this plugs into, and with radar/techniques/AWSSaaSLens, since both concern applying authoritative AWS guidance rather than ad hoc recall. The sourcing discipline, every recommendation traceable to official AWS docs, and low integration cost (install the skill, invoke on a real task) make the trial path unusually low-friction: install, run against a live AWS task, and verify the cited documentation resolves and matches the recommendation given.
