M1.5 accepted: tests over a published image #118

2026-10-07T23:09:09Z

dev-agent commented

2026-10-07 23:09:09 +00:00

Closes out M1.5. v0.1.0's published image passed the whole end-to-end suite: 9 of 9, in 11m 37s. The tests were the release's own, running over the image as published.

What changes

The Dockerfile's test stage stacks on TEST_BASE. By default that's this build's agent; it can also be a published image. It adds only the test tools (the dev group, through uv export, into the image's environment) and never reinstalls the agent. So what's under test is exactly what was published.
scripts/e2e.sh --image passes NISSE_E2E_MODEL into the container.
The plan marks M1.5 done and records how to run acceptance against a release.

How the acceptance was run

git worktree add /tmp/v010 v0.1.0          # the release's own tests
cp Dockerfile .dockerignore /tmp/v010/       # this PR's Dockerfile
docker build --target test --build-arg TEST_BASE=gitea.dongar.net/dev/nisse-dev:0.1.0 -t nisse-dev:0.1.0-test /tmp/v010
NISSE_E2E_MODEL=qwen3.8-27b-nothink scripts/e2e.sh --image nisse-dev:0.1.0-test

The run reported spending $0.0591 on the test key. That overlapped main's CI checks job (22:55–22:58), whose integration tier uses Haiku on the same key, so it's almost certainly that, not a paid call from the e2e run.

🤖 Generated with Claude Code

https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

Closes out M1.5. **`v0.1.0`'s published image passed the whole end-to-end suite: 9 of 9, in 11m 37s.** The tests were the release's own, running over the image as published. ## What changes - **The Dockerfile's `test` stage stacks on `TEST_BASE`.** By default that's this build's agent; it can also be a published image. It adds only the test tools (the dev group, through `uv export`, into the image's environment) and never reinstalls the agent. So what's under test is exactly what was published. - **`scripts/e2e.sh --image`** passes `NISSE_E2E_MODEL` into the container. - **The plan** marks M1.5 done and records how to run acceptance against a release. ## How the acceptance was run ``` git worktree add /tmp/v010 v0.1.0 # the release's own tests cp Dockerfile .dockerignore /tmp/v010/ # this PR's Dockerfile docker build --target test --build-arg TEST_BASE=gitea.dongar.net/dev/nisse-dev:0.1.0 -t nisse-dev:0.1.0-test /tmp/v010 NISSE_E2E_MODEL=qwen3.8-27b-nothink scripts/e2e.sh --image nisse-dev:0.1.0-test ``` The run reported spending $0.0591 on the test key. That overlapped `main`'s CI `checks` job (22:55–22:58), whose integration tier uses Haiku on the same key, so it's almost certainly that, not a paid call from the e2e run. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

dev-agent added 1 commit 2026-10-07 23:09:09 +00:00

M1.5 accepted: the tests run over a published image

CI / checks (push) Skipped

CI / checks (pull_request) Successful in 2m10s

Details

CI / e2e (pull_request) Successful in 2m54s

Details

d7e4fd0938

- The Dockerfile's test stage stacks on TEST_BASE: this build's agent by
  default, or a published image. It adds only the test tools (uv export
  of the dev group into the image's environment), never reinstalling the
  agent, so the suite tests the image as published.
- scripts/e2e.sh --image passes NISSE_E2E_MODEL into the container.
- The plan marks M1.5 done: v0.1.0's published image passed the whole e2e
  suite (9 of 9, 11m37s), with the release's own tests over it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PPQ14L3LikyefCMFBqEbKE

dongar approved these changes

2026-10-07 23:12:20 +00:00

dev-agent merged commit 572bd9f8eb into main

2026-10-07 23:14:45 +00:00

dev-agent deleted branch image-accept

2026-10-07 23:14:45 +00:00

Sign in to join this conversation.

2 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: dev/nisse#118