agentleFS
mejbaurbaharmejbaurbahar/software-tester-skills/skills/installation-upgrade-testing/SKILL.md

installation-upgrade-testing

Skill1 starsChanged 53 days agoMITView on GitHub

Use when testing software installation, upgrade, downgrade, uninstall and packaging — installers (MSI/DMG/DEB/RPM/Docker/Helm/npm/pip), silent installs, permission and disk edge cases, data/config preservation across versions, rollback, and version-skew.

  • Installs packages

The file

Files
SKILL.md

Installation, Upgrade & Uninstall Testing

First and last impressions of your product happen in the installer. Use clean VMs/containers with snapshots for every run.

Matrix

Fresh install · upgrade N-1→N (and N-2, LTS→latest) · skip-version upgrade · downgrade/rollback · repair/reinstall · uninstall/purge · side-by-side versions · per-user vs system-wide · air-gapped/offline · silent/unattended · interrupted install (kill, power loss) · re-run same installer (idempotent).

Checklist

AreaVerify
PrereqsMissing/old runtime (JRE/.NET/glibc/OpenSSL) → clear error; disk space check; OS/arch support (x64/arm64)
PermissionsNon-admin/non-root behavior, UAC/sudo prompts, SELinux/AppArmor, file ownership & modes (no world-writable)
PathsSpaces/unicode/long paths, custom install dir, network drive, read-only FS
IntegrityChecksums/signatures verified (sha256sum -c, gpg --verify, cosign verify), SBOM shipped, no debug symbols/secrets
Config & dataExisting config preserved (three-way merge), migrations run once & atomically, backup created before upgrade, user data untouched
ServicesDaemon/service registered, starts on boot, ports free/conflict handling, health check passes
UpgradeIn-flight jobs/connections drained; DB schema forward + backward compatible during rolling update; cache/session invalidation; rollback path tested
UninstallRemoves binaries, services, PATH/registry, shortcuts; keeps or prompts on user data; no orphan processes
Docs & UXInstall guide commands actually work copy-paste; progress, errors and logs actionable; exit codes (cli-testing)

Automation recipes

# Debian package in a clean container
docker run --rm -v $PWD:/pkg debian:12 bash -c 'apt-get update -qq && apt-get install -y /pkg/app_2.0.0_amd64.deb && app --version && apt-get purge -y app && test ! -e /usr/bin/app'
# Upgrade path matrix
for from in 1.8.0 1.9.3 2.0.0-rc1; do ./install.sh $from && seed_data && ./install.sh 2.0.0 && ./verify_data.sh; done
# Helm
helm lint chart/ && helm template chart/ | kubeconform -strict -summary
helm upgrade --install app chart/ --atomic --wait --timeout 5m && helm rollback app 1
# Windows silent
msiexec /i App.msi /qn /l*v install.log ; msiexec /x App.msi /qn
# npm/pip package sanity
npm pack && npm i -g ./pkg-1.0.0.tgz && pkg --version ; python -m venv v && v/bin/pip install dist/*.whl && v/bin/python -c "import pkg"

Docker image: non-root user, minimal base, healthcheck, reproducible tags, docker run with read-only FS, multi-arch manifests, image scan (iac-container-security-testing).

Failure injection

Kill installer at 10/50/90%, fill disk, revoke network mid-download, corrupt package, expire cert, pre-existing conflicting version, locked files/running app → expect atomic result: fully old or fully new, never half-installed.

Related

compatibility-testing, desktop-app-testing, data-migration-testing, cloud-infrastructure-testing, release-readiness-testing

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.