fix(docs-ci): probe the benchmark page at its published slug in the styles check - #13476
Conversation
…tyles check The RecipeStyles probe added in #12339 derived its URL from the page's file path (recipes/feature-benchmarks/llama-3-3-70b-topology), but the nav publishes the section as 'benchmarks' and the page under an explicit 'llama-3-70b-topology' slug, so the probe 404ed on every run. Every main-publish Fern Docs run that executed the check since Aug 4 failed on it; runs that looked green had skipped the docs job entirely. Verified against the live site: all six probes pass, including .dynamo-benchmark-grid (6 rules) at the corrected URL. Signed-off-by: Dan Gil <dagil@nvidia.com>
|
/ok to test bd599fd |
github-actions
Bot
added
fix
documentation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
Included review availability: Your plan includes up to 12 reviews per rolling hour; 10 remain after this review. WalkthroughThe published style checker now checks the updated benchmark recipe path. It also simplifies disallowed-origin error-message construction and adds spacing after the module docstring. ChangesPublished style checker
Estimated code review effort: 1 (Trivial) | ~2 minutes Merge Risk: ⚪ Minimal · up to The styles check now probes the benchmark page at its published URL, correcting the documented 404 failure without changing product behavior; no actionable merge-blocking risk remains after normal checks and review. 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Comment |
![devin-ai-integration[bot]](https://avatars.githubusercontent.com/in/811515?s=60&v=4){kind=small}
![dynamo-review-agent[bot]](https://avatars.githubusercontent.com/u/201626793?s=60&v=4){kind=small}
BenHamm
left a comment
BenHamm
left a comment
There was a problem hiding this comment.
Approving. Verified the diagnosis against the live site rather than the description.
- The old probe URL,
/dynamo/dev/recipes/feature-benchmarks/llama-3-3-70b-topology, returns 404. - The new one,
/dynamo/dev/recipes/benchmarks/llama-3-70b-topology, returns 200.
So the file-path-derived URL never matched the published slug, and the step could only ever fail. The nav confirms why: the page carries an explicit slug: llama-3-70b-topology distinct from its filename, under a section published as benchmarks rather than the feature-benchmarks directory name.
Worth noting the fix leaves the check meaningful rather than merely green: the page at the corrected URL serves 126 matches for the dynamo-* component classes, so the probe will genuinely assert that the component CSS shipped. A probe that 404s proves nothing; this one now proves the thing it was written to prove.
The second hunk is an unrelated f-string concatenation tidy-up in the same file. Harmless.
CI clean, no unresolved threads. Good catch on the two weeks of masked failures.
2 participants
Summary
Every main-publish run of the Fern Docs workflow has failed its final step, "Verify published pages carry their component CSS", since the RecipeStyles probe was added on Aug 4 (#12339). The probe derived its URL from the page's file path (
recipes/feature-benchmarks/llama-3-3-70b-topology), but the nav publishes the section underslug: benchmarksand the page under an explicitslug: llama-3-70b-topology, so the probe 404ed on every execution. Runs that looked green had merely skipped the docs job (non-docs pushes). The failure was masked for two weeks because the step runs after the sync and publish already succeeded.One-line fix: probe
recipes/benchmarks/llama-3-70b-topology, with a comment noting the URL comes from the nav's explicit slugs, not the file path.Validation
Ran
check_published_styles.pyagainst the live site: all six probes pass, including.dynamo-benchmark-grid(6 rules, RecipeStyles) at the corrected URL. The first main publish after this merges should go fully green end to end.🤖 Generated with Claude Code
Summary by CodeRabbit