feat(evals): add missing docker-via-wsl eval#41
Merged
Conversation
The 24 existing evals cover docker-development only; docker-via-wsl (added in a later PR) had none. Adds one eval asserting the skill's core correction: re-issue the failing Docker command through `wsl.exe -e bash -lc` from inside WSL2's native filesystem, instead of continuing to drive Docker from a Windows/Git-Bash shell on a mapped network drive (the actual cause of the "Is a directory" bind- mount failure). Signed-off-by: Sebastian Mendel <info@sebastianmendel.de>
Contributor
Dependency Review✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.Scanned FilesNone |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request adds a new evaluation test case, "docker-via-wsl-reissue-not-windows-shell", to "evals/evals.json". The reviewer pointed out a logical issue in one of the regular expression assertions, where "-e bash -lc|bash -lc" simplifies to just "bash -lc" and would incorrectly allow invocations without the "-e" flag to pass. A suggestion was provided to simplify the regex to "-e bash -lc".
Important
The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.
The alternation '-e bash -lc|bash -lc' was redundant since 'bash -lc' matches as a substring of the first alternative, letting a response that omits the -e flag pass the discriminating assertion. Signed-off-by: Sebastian Mendel <info@sebastianmendel.de>
|
CybotTM
added a commit
to netresearch/skill-repo-skill
that referenced
this pull request
Jul 12, 2026
…163) Part of #148 (A6: eval backfill + quality rule). ## Summary Adds a quality rule to `references/materialization-contract.md` Rule 7 (Eval format): every eval needs at least one assertion a no-skill baseline answer FAILS, verified locally with `run-ab-evals.sh` (the `without_skill` pass rate on that assertion must be <1.0). Keyword assertions any generic answer already satisfies (`"alpine"`, `"USER"`, a bare `"Plugin"`) reward boilerplate and never test the retro-born gotcha that is a skill's actual value. The rule explicitly notes this verification is a **local/manual step, not CI** — `run-ab-evals.sh` shells out to `claude -p` per eval and is intentionally not wired into any workflow. ## Companion PRs (backfilled per this rule) - netresearch/jujutsu-workflow-skill#10 - netresearch/markdown-to-pdf-skill#7 - netresearch/typo3-upgrade-effort-model-skill#14 - netresearch/typo3-site-conformance-skill#9 - netresearch/docker-development-skill#41 ## Validation ``` $ npx markdownlint-cli2 skills/skill-repo/references/materialization-contract.md Summary: 0 error(s) ``` Pre-commit hooks (trim-trailing-whitespace, end-of-file-fixer, markdownlint-cli2) passed on commit.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Part of netresearch/skill-repo-skill#148 (A6: eval backfill).
Summary
The 24 existing evals in
evals/evals.jsoncover thedocker-developmentskill only;docker-via-wsl(added in a later PR) had none. Adds one eval,docker-via-wsl-reissue-not-windows-shell.Delta-discriminating assertion
Asserts the skill's core correction for an agent whose Bash shell is Git Bash/MSYS on Windows outside WSL: re-issue the failing Docker command through
wsl.exe -e bash -lcfrom inside WSL2's native filesystem, instead of continuing to drive Docker from the Windows shell on a mapped network drive (the actual root cause of the "Is a directory" bind-mount failure, perskills/docker-via-wsl/SKILL.md). A generic baseline diagnoses this as a Docker cache bug or a fragile single-file-bind-mount problem and has no way to know thewsl.exe -e bash -lcre-issue pattern without having read this skill.Validation