Removes the .gitea line from .dockerignore. The development image already received .git but not .gitea, so in /src git saw the CI workflow as deleted: git status was not empty, make build stamped a -dirty version, and git commit -a would have deleted the workflow. With the workflow in the build context, /src matches the commit the image was built from.
Unmet definition-of-done item: .dockerignore already differed from the sneak/promptsnext template before this change, because the template changed on 2026-10-04 (submodule .git/config patterns, hardware-backed SSH key files). Bringing it up to date is outside this issue.
Model: opus-5-5
Closes https://git.eeqj.de/sneak/keyfunc/issues/63.
Removes the `.gitea` line from `.dockerignore`. The development image already received `.git` but not `.gitea`, so in `/src` git saw the CI workflow as deleted: `git status` was not empty, `make build` stamped a `-dirty` version, and `git commit -a` would have deleted the workflow. With the workflow in the build context, `/src` matches the commit the image was built from.
- Unmet definition-of-done item: `.dockerignore` already differed from the `sneak/prompts` `next` template before this change, because the template changed on 2026-10-04 (submodule `.git/config` patterns, hardware-backed SSH key files). Bringing it up to date is outside this issue.
Model: opus-5-5
.dockerignore, the comment above /keyfunc (lines 65–66) still says the CI workflow is kept out of the build context ("and the CI workflow, which is not a build input"). After this change the workflow is in the build context, so the comment now states the opposite of what the file does and invites a reader to put the .gitea line back. That clause belonged to the removed line. Acceptable: in the same commit the comment reads only # The binary `make build` writes. (the wording .gitignore uses), nothing else in the file changes, and the PR body drops its judgement-call line about this comment.
Model: opus-5-5
`.dockerignore`, the comment above `/keyfunc` (lines 65–66) still says the CI workflow is kept out of the build context ("and the CI workflow, which is not a build input"). After this change the workflow is in the build context, so the comment now states the opposite of what the file does and invites a reader to put the `.gitea` line back. That clause belonged to the removed line. Acceptable: in the same commit the comment reads only ``# The binary `make build` writes.`` (the wording `.gitignore` uses), nothing else in the file changes, and the PR body drops its judgement-call line about this comment.
Model: opus-5-5
The development image copies .git into /src but left .gitea out, so in
the image git saw the CI workflow as deleted: `git status` was not
empty, `make build` stamped a -dirty version, and `git commit -a` would
have deleted the workflow. With the .gitea line gone from
.dockerignore, /src matches the commit it was built from. The comment
above /keyfunc now names only the binary, as .gitignore does.
Model: opus-5-5
The comment above /keyfunc in .dockerignore now reads only # The binary `make build` writes., in the same commit, and the PR body no longer carries the judgement-call line about it.
Model: opus-5-5
The comment above `/keyfunc` in `.dockerignore` now reads only ``# The binary `make build` writes.``, in the same commit, and the PR body no longer carries the judgement-call line about it.
Model: opus-5-5
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Closes #63.
Removes the
.gitealine from.dockerignore. The development image already received.gitbut not.gitea, so in/srcgit saw the CI workflow as deleted:git statuswas not empty,make buildstamped a-dirtyversion, andgit commit -awould have deleted the workflow. With the workflow in the build context,/srcmatches the commit the image was built from..dockerignorealready differed from thesneak/promptsnexttemplate before this change, because the template changed on 2026-10-04 (submodule.git/configpatterns, hardware-backed SSH key files). Bringing it up to date is outside this issue.Model: opus-5-5
.dockerignore, the comment above/keyfunc(lines 65–66) still says the CI workflow is kept out of the build context ("and the CI workflow, which is not a build input"). After this change the workflow is in the build context, so the comment now states the opposite of what the file does and invites a reader to put the.gitealine back. That clause belonged to the removed line. Acceptable: in the same commit the comment reads only# The binary `make build` writes.(the wording.gitignoreuses), nothing else in the file changes, and the PR body drops its judgement-call line about this comment.Model: opus-5-5
835e6a6992toc26bf98033The comment above
/keyfuncin.dockerignorenow reads only# The binary `make build` writes., in the same commit, and the PR body no longer carries the judgement-call line about it.Model: opus-5-5
Review passed.
Model: opus-5-5