skywalkskills

CataloguePerformance, Scale & Cost

C3Ready

Cloud Cost & Right-Sizing Audit

Prices the AWS bill line by line: idle resources, oversized compute, storage left on an expensive tier, and traffic that takes a costly path.

What it examines

  • Stopped or unused compute, unattached addresses, and idle load balancers
  • Instances larger than their usage supports
  • Storage with no lifecycle to a cheaper tier
  • Unattached volumes and old snapshots
  • NAT gateway traffic a VPC endpoint would carry
  • Cross-AZ transfer that placement can avoid
  • Steady On-Demand compute with no Savings Plan

What you get back

A findings file. Each item has the evidence, how serious it is, and a fix.

Dependencies

  • An AI coding agent that reads SKILL.md files, such as Claude Code, Codex or Gemini CLI
  • An AWS account to audit. Live mode needs credentials, and the audit never uses the AWS CLI.
  • AWS credentials that boto3 can find: environment variables, a profile selected with AWS_PROFILE, or SSO. Attach only the read policy in iam-policy.json. It lists the exact read-only permissions and no write access.
  • Cost Explorer turned on in the account. Compute Optimizer must be opted in for the right-sizing checks.
  • Python 3.12 or newer. The tool installs boto3 and jsonschema.
  • Python 3 with the jsonschema package, to check findings.json before fixing anything

How it works

The skill tells the agent to run the bundled c3-cost tool, not to improvise AWS commands. The tool reads billing and resource data and writes findings.json to standard output.

It runs seven checks: idle and orphaned resources, over-provisioned compute, wrong storage tier, orphan volumes and old snapshots, NAT traffic a VPC endpoint could carry, avoidable cross-AZ traffic and missing Savings Plans. Amounts are in rand.

A fixture mode runs the tool on sample data with no AWS access, for demos and checks. The remediate skill reads findings.json and writes staged plans, never changes made directly in AWS.

Setup

  1. Unzip the download and keep the whole folder. Copy its audit and remediate folders into your agent's skills folder (for Claude Code, .claude/skills/) in the repository you want to check.
  2. Create an IAM policy from iam-policy.json and attach it to the user or role you will sign in with. Make sure Cost Explorer is on and Compute Optimizer is opted in.
  3. Sign in to AWS: run aws sso login, or set AWS_PROFILE or the access key variables. Confirm with aws sts get-caller-identity.
  4. Install the tool: pip install boto3 jsonschema. To try it without AWS first, run the tool with --fixture.
  5. Open your agent and ask it to run the c3-cloud-cost-audit skill. It runs the tool in live mode and saves findings.json.
  6. Check the result with the validator: pip install jsonschema, then python3 path/to/the-unzipped-folder/schema/validate.py findings.json.
  7. Ask your agent to run the c3-cloud-cost-remediate skill using findings.json. Fix some findings and leave the rest open if you like. Each change needs your approval.

Run it with your agent

Copy the prompt, fill in the parts in square brackets, and paste it into your agent.

Audit prompt

Run the c3-cloud-cost-audit skill. Use the AWS profile [profile name]. Run the bundled c3-cost tool in live mode and do not use the AWS CLI. Save its output as findings.json, then validate it with schema/validate.py. To try it first without AWS access, run the tool with --fixture.

Fix prompt, after you have findings.json

Run the c3-cloud-cost-remediate skill using findings.json. Work through the findings I choose, and leave the rest open.

Seen on ClaimsPortalA bill that does not match the machines you think you runRead the finding

What is inside

The files that make up this skill, as your agent reads them.

packages/c3-cloud-cost

audit › SKILL.md

Sign in to download

Use the email and password you signed up with.