AWS Price CSV Skill

Overview

Transforms a user-provided AWS service list (instances, volumes, S3 buckets, etc.) into a pricing CSV. The script can query the AWS Price List API (via aws-cli) or reuse cached bulk JSON files. It supports On-Demand and Reserved terms and automatically adds up per-item and total costs.

Quick Start

1. Prepare a YAML/JSON file with name, service_code, filters, term, and usage fields (see sample in references/api_reference.md).
2. Pick the data source:
   • API mode – requires aws pricing get-products permission and an internet connection.
   • Bulk mode – no IAM access required; the script downloads/caches public bulk JSON files.
3. Run the script with the region and desired options:

   python3 scripts/generate_pricing_csv.py \
     --input inputs/sample.yml \
     --region ap-northeast-1 \
     --source bulk \
     --cache-dir ~/.cache/aws-price-csv \
     --output quotes/apac_quote.csv
   

4. Inspect the CSV (each line item + TOTAL) and deliver alongside the original request if needed.

Workflow

1. Prepare the input list

• Use the YAML/JSON templates in references/api_reference.md.
• filters must map to AWS pricing attributes (instanceType, regionCode, volumeApiName, usagetype, termType, etc.).
• term.type: OnDemand or Reserved (RI).
• term.attributes (optional) filter Reserved prices (LeaseContractLength, PurchaseOption, OfferingClass, ...).
• usage.quantity represents the amount to multiply (hours, GB-Mo, requests, ...).

2. Choose the data source

| Mode | When to use | Notes | |------|-------------|-------| | API (--source api) | You already have IAM creds and want real-time data | Uses aws pricing get-products in us-east-1 | | Bulk (--source bulk) | Offline, no IAM, or you want caching | The script checks --cache-dir (default ~/.cache/aws-price-csv); cached files newer than 30 days are reused, otherwise they are re-downloaded |

You can still override with --bulk-files ServiceCode=/path/to/file.json; those files win over the cache.

3. Generate the CSV

• Script path: scripts/generate_pricing_csv.py
• Key arguments:
  • --input: YAML/JSON list
  • --region: AWS region code (location name is auto-added to filters)
  • --output: CSV path (default aws_pricing.csv)
  • --source: api (default) or bulk
  • --cache-dir: bulk cache directory (default ~/.cache/aws-price-csv)
  • --force-refresh: ignore cached bulk files and re-download
  • --bulk-files: ServiceCode=/path/to/file.json overrides the cache
• Output columns: item_name, service_code, term_type, region, location, quantity, usage_unit, price_unit, price_per_unit_usd, cost_usd, description, plus a TOTAL row.

4. Validate & deliver

• Review the CSV to ensure every line item has pricing and a description.
• Confirm the TOTAL matches expectations.
• Include the source list or any supporting docs if required.

Troubleshooting

| Issue | Fix | |-------|-----| | aws pricing returns empty results | Double-check filters (location, regionCode, termType, etc.) or fall back to bulk mode | | aws CLI missing | Install aws-cli v2, configure credentials, or rely on bulk mode | | Need to refresh cached data | Use --force-refresh or delete the cached JSON so it gets re-downloaded | | PyYAML missing | pip install pyyaml or convert input to JSON | | Reserved price not found | Add LeaseContractLength, PurchaseOption, OfferingClass inside term.attributes |

Resources

• scripts/generate_pricing_csv.py – main script with API/bulk support, caching logic, and On-Demand/Reserved handling.
• references/api_reference.md – Price List API examples, bulk download pointers, region/location table, and sample input templates (including gp3 and RI cases).