Repository navigation
test: add AILF tasks and report - #3294
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub. 4 Skipped Deployments
|
|
somerandomdude
left a comment
There was a problem hiding this comment.
LGTM as a good first pass. We'll no doubt be refining these in future iterations. Thanks again!
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 2 potential issues.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Want fixes drafted automatically? Bugbot Autofix can create code changes for findings. A team admin can enable Autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 6ad7a30. Configure here.
| - type: contains | ||
| value: '@sanity/icons/Search' | ||
| - type: contains | ||
| value: 'icon={SearchIcon}' |
There was a problem hiding this comment.
Icon task rejects valid element form
Medium Severity
The Icon wrapper rubric accepts both icon={SearchIcon} and icon={<SearchIcon />}, but the contains assertion requires only the component form. A correct answer using the documented element form fails the string check.
Reviewed by Cursor Bugbot for commit 6ad7a30. Configure here.
There was a problem hiding this comment.
We'll address this later


Description
This PR moves our Sanity UI 5 AILF tasks from the AILF studio into the repo for better tracking/iteration along with the components. A copy of the report from an initial run with these tasks is also included.
Once we land v5 in the
mainbranch, we can turn this into a GitHub workflow that we can trigger from the repo, or PRs. In the mean time, it needs to be run from the repo root in a terminal withnpx @sanity/ailf@latest run --remote. A--debugflag can be added if you want to test things out.What to review
Are the tasks and config as expected?
Testing
Ran locally to verify (which generated the report included in this PR).
Note
Low Risk
Adds evaluation config, tasks, and a report snapshot only; no changes to
@sanity/uiruntime or published packages.Overview
Adds the AI Literacy Framework (AILF) setup under
.ailf/so Sanity UI v5 can be evaluated from this repo instead of only from the AILF studio.ailf.config.tspoints at production docs, assigns ownership to the SDK team, and loads tasks from the repo.The bulk of the change is literacy task definitions (
.ailf/tasks/*.yaml) and matching canonical reference solutions for migration and API pitfalls—v3→v5 props, tone/levelvocabulary, icon subpaths, polyfills, accessibility patterns, and similar. A snapshot of the latest remote run is committed in.ailf/results/latest/(report shows 81/100 onsanity-ui-v5, 516 tests).Supporting tweaks:
.env.exampledocumentsAILF_API_KEYfornpx @sanity/ailf run --remote,.gitignoreignores.env, and oxfmt/oxlint skip.ailf/**so generated YAML/TSX does not get formatted or linted like app code.Reviewed by Cursor Bugbot for commit 5965ee2. Bugbot is set up for automated code reviews on this repo. Configure here.