## Thinking Path
> - Paperclip is the open source control plane people use to manage
AI-agent companies
> - Managed updates can change both the application payload and its
database schema
> - Unit tests cannot prove that an older live install upgrades through
a real migration and remains recoverable
> - The managed-install work in #10045 needs a repeatable cross-version
system test
> - This pull request adds an isolated end-to-end harness for update,
migration, backup, service restart, and rollback behavior
> - The benefit is a direct proof that managed upgrades preserve the
existing database and service lifecycle across versions
## Linked Issues or Issue Description
Refs #10045
This test is a focused follow-up to the managed install integration.
Merge #10045 first so the tested install, update, service, backup, and
rollback commands are available.
## What Changed
- Added a cross-version managed-update E2E script.
- Installed an older Git ref, initialized its embedded PostgreSQL
database, and updated to a ref with one additional migration.
- Verified the pre-update backup, payload switch, service recovery,
migration result, database-cluster reuse, and rollback behavior.
- Isolated Paperclip state under a dedicated test home and cleaned up
the service and managed install on success or failure.
- Added regression tests for shell syntax, required-ref validation,
side-effect-free preflight failure, and complete failure cleanup.
## Verification
- `node --test scripts/__tests__/e2e-update-migrations.test.mjs`
- `bash -n scripts/e2e-update-migrations.sh`
- GitHub latest-head CI: build, typecheck, release registry, canary
dry-run, general tests, serialized suites, and both browser E2E shards
passed.
- Full harness execution needs an isolated macOS or Linux host with a
real launchd or systemd user service. It is intentionally not run on a
live Paperclip server host.
## Risks
- The script manages a real user service and downloads two Git refs. Run
it only on an isolated test host.
- The test needs #10045 because `origin/master` does not yet contain the
managed install lifecycle.
- The script uses a dedicated `PAPERCLIP_HOME`, refuses a pre-existing
shim or test home, and removes its service and install during cleanup.
> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.
## Model Used
- OpenAI Codex based on GPT-5. The runtime did not expose a more
specific deployment ID or context-window size. Reasoning, repository
access, shell execution, and GitHub tooling were enabled.
## Checklist
- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] All Paperclip CI gates are green
- [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge
---------
Co-authored-by: Paperclip <noreply@paperclip.ing>
Dotta