v0.2.0

0.2.0 (2026-08-26)

⚠ BREAKING CHANGES

• the scan reports failed regions and services in its output and exit code (#69)
• scan documents move out of shared /tmp into a per-user directory (#66)
• remove --format — the scan always writes JSON, with --output - for stdout (#61)
• ECS service ids are the cluster/service path AWS puts in the ARN (#65)
• scan output is a self-describing JSON document (#58)
• resource names are real AWS names or null (#57)
• resource types always start with the CLI service key (#56)
• every resource has a real ID and a real ARN (#55)

✨ Features

• default output filenames start with the AWS account id (#79) (da1566a)
• every resource has a real ID and a real ARN (#55) (2f475e6)
• remove --format — the scan always writes JSON, with --output - for stdout (#61) (e9518ca)
• resource names are real AWS names or null (#57) (ac3c9cd)
• resource types always start with the CLI service key (#56) (de60f07)
• scan output is a self-describing JSON document (#58) (82ddbac)
• the results table shows names and ids under an account header (#59) (1c6972a)
• the scan reports failed regions and services in its output and exit code (#69) (aeaf677)

🐞 Bug Fixes

• e2e-diff.sh stops passing the --format flag that was removed (#67) (89bf221)
• ECS service ids are the cluster/service path AWS puts in the ARN (#65) (b1eb220)
• every scanning client gets its own botocore config (#60) (2710c55)
• logging is configured only by the CLI, never as an import side effect (#68) (ac19cb9)
• scan documents move out of shared /tmp into a per-user directory (#66) (d0fb28f)
• the scan cache moves out of shared /tmp into a per-user directory (#62) (e550a27)

📦 Other Changes

• ADR-0005 records the resource identity and output model (#53) (957bd1e)
• delete the completed Poetry migration guide (#49) (50aa2ce)
• one definition of the scan-path choice (#52) (65d9ddf)
• records.py states the name guarantee the code actually gives (#64) (724b4f0)

v0.1.1

0.1.1 (2026-08-22)

Important

0.1.0 was published with a broken wheel. cli.py was left out of the distribution, so the installed aws-inventory command died immediately with ModuleNotFoundError: No module named 'cli'. 0.1.1 is the first working release; 0.1.0 is yanked on PyPI.

uv tool install aws-resource-inventory   # or: pipx install aws-resource-inventory
aws-inventory --help

Module paths also moved under a single aws_resource_inventory/ package. That only affects direct imports, and since 0.1.0 never ran, nothing depended on the old paths.

🐞 Bug Fixes

• e2e-diff refuses refs it cannot honestly compare (#48) (215d844)
• installed CLI starts — ship cli.py and aws_scanner.py in the wheel (#45) (c926547)
• release pipeline publishes to PyPI, and packaging is checked before release (#42) (b4d7935)

📦 Other Changes

• everything ships inside one aws_resource_inventory package (#47) (2f284e9)
• install from PyPI with uv or pipx (#46) (b80f403)

v0.1.0

0.1.0 (2026-08-22)

Initial release of aws-resource-inventory — a read-only CLI (aws-inventory) that inventories AWS resources across regions and services.

✨ Highlights

• Scans eight AWS services — EC2, S3, ECS, EFS, VPC, RDS, ELB, and Auto Scaling — across multiple regions concurrently, with or without tag filters.
• Discovers resources from 100+ AWS services via the Resource Groups Tagging API (--all-services with a tag filter).
• Table, JSON, and Markdown output — rendered in the terminal and written to a file for further processing.
• Result caching with a 10-minute TTL, configurable parallelism per region (--max-workers) and per service (--service-workers), dry-run preview, continuous refresh mode, and graceful Ctrl+C handling.
• Debug traces (--debug), full AWS API tracing (--verbose), custom log files (--log-file), and shell tab-completion for commands, options, and service names.

Install: pip install aws-resource-inventory