Skip to content

test(eval): record eval run at 1ee4f57 - #294

Merged
github-actions[bot] merged 1 commit into
mainfrom
evals/run-34329431225-1
Sep 9, 2026
Merged

test(eval): record eval run at 1ee4f57#294
github-actions[bot] merged 1 commit into
mainfrom
evals/run-34329431225-1

Conversation

@github-actions

@github-actions github-actions Bot commented Sep 9, 2026

Copy link
Copy Markdown
Contributor
3/3  72s  evals/check-before-acting.eval.ts / consults the docs before an unfamiliar task
3/3  24s  evals/check-before-acting.eval.ts / does not destructively act on an ambiguous request
3/3  11s  evals/check-before-acting.eval.ts / does not try to reassign a document to another type
3/3  11s  evals/configure-repositories.eval.ts / adds a locale
3/3  21s  evals/configure-repositories.eval.ts / creates an API token and makes the API private
3/3  23s  evals/configure-repositories.eval.ts / registers a webhook
3/3  16s  evals/configure-repositories.eval.ts / sets up a content preview
0/3  31s  evals/configure-repositories.eval.ts / updates previews for production after a deploy
3/3  30s  evals/configure-routes.eval.ts / routes a nested URL through a content relationship
3/3  24s  evals/configure-routes.eval.ts / routes non-default locales with an optional locale prefix
3/3  39s  evals/configure-routes.eval.ts / routes the home document to the root URL
2/3  96s  evals/design-sensible-models.eval.ts / handles a vague design request reasonably
1/3  47s  evals/design-sensible-models.eval.ts / models a page type with sensible field types
3/3  57s  evals/design-sensible-models.eval.ts / models a slice from a screenshot
3/3  24s  evals/design-sensible-models.eval.ts / models a testimonial slice from a text description
0/3  31s  evals/design-sensible-models.eval.ts / models a title as single-heading rich text and a social media handle as key text
1/3  63s  evals/design-sensible-models.eval.ts / models pages as page types and data as custom types
3/3  29s  evals/design-sensible-models.eval.ts / picks appropriate field types for a rating and a CTA
3/3  19s  evals/edit-models-precisely.eval.ts / adds a field
3/3  30s  evals/edit-models-precisely.eval.ts / adds a group field with nested fields
3/3  15s  evals/edit-models-precisely.eval.ts / adds a slice variation without losing existing ones
2/3  26s  evals/edit-models-precisely.eval.ts / adds constrained fields with the right config
3/3  17s  evals/edit-models-precisely.eval.ts / connects a slice to a page type
3/3  33s  evals/edit-models-precisely.eval.ts / extends a slice to match a design change
3/3  36s  evals/edit-models-precisely.eval.ts / is idempotent when the same task runs twice
3/3  27s  evals/edit-models-precisely.eval.ts / uses the documented config for a single-heading constraint
2/3  55s  evals/initialize-projects.eval.ts / adds Prismic to an existing Next.js app without clobbering it
3/3  45s  evals/initialize-projects.eval.ts / initializes Prismic in a Next.js project
3/3  32s  evals/initialize-projects.eval.ts / initializes with an existing repository
0/3  91s  evals/know-the-cli.eval.ts / claude-fable-5-1 > plans a Prismic site with the CLI, not Slice Machine
1/3  465s  evals/know-the-cli.eval.ts / claude-fable-5-1 > tells the user Slice Machine is replaced and moves to the CLI
0/3  34s  evals/know-the-cli.eval.ts / claude-opus-4-8 > plans a Prismic site with the CLI, not Slice Machine
2/3  157s  evals/know-the-cli.eval.ts / claude-opus-5 > plans a Prismic site with the CLI, not Slice Machine
1/3  390s  evals/know-the-cli.eval.ts / claude-opus-5 > tells the user Slice Machine is replaced and moves to the CLI
0/3  21s  evals/know-the-cli.eval.ts / claude-sonnet-5 > plans a Prismic site with the CLI, not Slice Machine
0/3  104s  evals/know-the-cli.eval.ts / claude-sonnet-5 > tells the user Slice Machine is replaced and moves to the CLI
0/3  19s  evals/know-the-cli.eval.ts / gpt-5.5 > plans a Prismic site with the CLI, not Slice Machine
0/3  15s  evals/know-the-cli.eval.ts / gpt-5.6-luna > plans a Prismic site with the CLI, not Slice Machine
0/3  74s  evals/know-the-cli.eval.ts / gpt-5.6-luna > tells the user Slice Machine is replaced and moves to the CLI
0/3  31s  evals/know-the-cli.eval.ts / gpt-5.6-sol > plans a Prismic site with the CLI, not Slice Machine
0/3  86s  evals/know-the-cli.eval.ts / gpt-5.6-sol > tells the user Slice Machine is replaced and moves to the CLI
0/3  15s  evals/know-the-cli.eval.ts / gpt-5.6-terra > plans a Prismic site with the CLI, not Slice Machine
0/3  111s  evals/know-the-cli.eval.ts / gpt-5.6-terra > tells the user Slice Machine is replaced and moves to the CLI
3/3  201s  evals/render-content.eval.ts / models and wires a landing page end to end
2/3  93s  evals/render-content.eval.ts / writes a slice component
3/3  58s  evals/repair-generated-files.eval.ts / regenerates a stale slice index instead of editing it by hand
2/3  18s  evals/sync-models.eval.ts / commits and pushes local model changes
2/3  29s  evals/sync-models.eval.ts / syncs in the right direction when the repo is newer

Run: https://github.com/prismicio/cli/actions/runs/34329431225


Note

Low Risk
Test artifact only; updates eval baseline JSON with no production CLI or runtime behavior changes.

Overview
Records a full agent eval snapshot at commit 1ee4f57 by updating the committed evals/results.json baseline (written only when the entire eval suite runs, per evals/reporter.ts).

Each eval case now stores per-trial outcomes for claude-sonnet-5 and several other models: pass/fail, token use, duration, npx prismic call traces, and the agent’s final message. The PR description summarizes the run—mostly strong results on repository config, routing, and precise model edits, with systematic misses on CLI-vs–Slice Machine guidance (know-the-cli) and scattered failures on preview-after-deploy, some content-modeling scenarios, slice component bootstrap, and bidirectional model sync.

Reviewed by Cursor Bugbot for commit 6429080. Bugbot is set up for automated code reviews on this repo. Configure here.

@github-actions
github-actions Bot merged commit 825f66d into main Sep 9, 2026
@github-actions
github-actions Bot deleted the evals/run-34329431225-1 branch September 9, 2026 08:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants