test(eval): align the Slice Machine criteria with the deprecation - #308
Merged
Merged
Conversation
The Slice Machine init halt has no flag to continue. The redirect criterion now says init exits with code 1 on every run and fails an agent that works around the halt without telling the user. The knowledge criterion accepts the Type Builder as valid tooling and fails a plan that runs Slice Machine init or start. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0154ki8C9BX9VcoQFgARKJFp
lihbr
approved these changes
Sep 17, 2026
lihbr
left a comment
lihbr
left a comment
Member
There was a problem hiding this comment.
Looking good, to be settled with: prismicio/slice-machine#1779 (comment)
The deprecated Slice Machine init now continues when the user passes `--force`. An agent must not choose that flag on its own, so the judge fails a run that passes it without the user. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0154ki8C9BX9VcoQFgARKJFp
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Want reviews to match your repository better? Bugbot Learning can learn team-specific rules from PR activity. A team admin can enable Learning in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 4ae8142. Configure here.
The planning criteria only accepted the CLI, so a plan that models content in the Type Builder failed, and the fail line contradicted the line below it about mentioning Slice Machine as deprecated. The redirect criteria now name the ways around the halt that should fail: pinning an older init version, or running start-slicemachine. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0154ki8C9BX9VcoQFgARKJFp
angeloashmore
changed the title
The planning eval grades a plan, not commands, so a plan that proposes Slice Machine fails whether or not it runs anything. The fail line goes back to describing the plan, and the line below it still allows mentioning Slice Machine as deprecated. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_0154ki8C9BX9VcoQFgARKJFp
angeloashmore
changed the title
angeloashmore
deleted the
claude/slice-machine-redirect-eval-criteria
branch
September 17, 2026 22:34
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
3 participants
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.
Resolves: Slice Machine Deprecation Steps
Description
Before this PR, the Slice Machine criteria did not describe how Prismic works today, so an agent could do the right thing and still fail.
After this PR, the criteria match the deprecation. Modeling content in the Type Builder counts as the current workflow, and reaching Slice Machine by another route counts as avoiding the Prismic CLI.
Checklist
Preview
How to QA 1
Read the two judge criteria in
evals/know-the-cli.eval.ts.🤖 Generated with Claude Code
https://claude.ai/code/session_0154ki8C9BX9VcoQFgARKJFp
Footnotes
Please use these labels when submitting a review:
⚠️ #issue: Strongly suggest a change.
❓ #ask: Ask a question.
💡 #idea: Suggest an idea.
🎉 #nice: Share a compliment. ↩