marketing: refresh compare pages and add compare update workflow#2699
Merged
Kitenite merged 1 commit intoMar 21, 2026
Merged
Conversation
Contributor
📝 WalkthroughWalkthroughThis PR refreshes Superset's marketing comparison pages by introducing an agent workflow command, adding three new comparison/tutorial pages, updating ten existing comparison pages with repositioned product descriptions, removing legacy planning documentation, and enhancing supporting code utilities for dynamic page type labeling, FAQ extraction, and JSON-LD metadata generation. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~22 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (8)
apps/marketing/src/components/JsonLd/JsonLd.tsx (1)
140-142: Normalize keyword values before adding them to JSON-LD.
keywordsis passed through as-is. Trimming and dropping empty entries avoids low-quality schema payloads from frontmatter drift.Proposed refactor
export function ComparisonJsonLd({ title, description, publishedTime, modifiedTime, url, image, keywords, articleSection, }: ComparisonJsonLdProps) { + const normalizedKeywords = Array.from( + new Set((keywords ?? []).map((k) => k.trim()).filter((k) => k.length > 0)), + ); + const schema = { "@context": "https://schema.org", "@type": "Article", headline: title, description, ...(articleSection && { articleSection }), - ...(keywords && keywords.length > 0 && { keywords }), + ...(normalizedKeywords.length > 0 && { keywords: normalizedKeywords }),🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/src/components/JsonLd/JsonLd.tsx` around lines 140 - 142, Normalize the keywords array before injecting into the JSON-LD object: transform the existing keywords variable by mapping each entry to its trimmed string and filtering out empty/falsy results (e.g., keywords = keywords.map(k => k.trim()).filter(Boolean)), then only spread into the payload using the existing conditional spread (the place that currently does ...(keywords && keywords.length > 0 && { keywords })); update the code around the JsonLd.tsx component where keywords is used so the JSON-LD contains only non-empty, trimmed keyword values.apps/marketing/src/app/compare/page.tsx (1)
33-35: Optional: drive section rendering from a small config array.Now that there are 3 page types, a config-based render loop will make future type additions less repetitive.
Refactor sketch
const oneVsOne = pages.filter((p) => p.type === "1v1"); const roundups = pages.filter((p) => p.type === "roundup"); const tutorials = pages.filter((p) => p.type === "tutorial"); + const sections = [ + { title: "Roundups", pages: roundups }, + { title: "Workflow Tutorials", pages: tutorials }, + { title: "Head-to-Head Comparisons", pages: oneVsOne }, + ].filter((section) => section.pages.length > 0); ... - {roundups.length > 0 && ( - <CompareSection title="Roundups" pages={roundups} /> - )} - - {tutorials.length > 0 && ( - <CompareSection title="Workflow Tutorials" pages={tutorials} /> - )} - - {oneVsOne.length > 0 && ( - <CompareSection title="Head-to-Head Comparisons" pages={oneVsOne} /> - )} + {sections.map((section) => ( + <CompareSection + key={section.title} + title={section.title} + pages={section.pages} + /> + ))}Also applies to: 73-83
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/src/app/compare/page.tsx` around lines 33 - 35, The code repeats pages.filter calls and section JSX for each type (oneVsOne, roundups, tutorials); replace these with a small config array like [{ key: "1v1", title: "1v1", id: "one-v-one" }, ...] and map over it to compute filteredPages = pages.filter(p => p.type === cfg.key) and render each section (use the same rendering logic currently in the page sections around the existing oneVsOne/roundups/tutorials blocks). Update any local identifiers (oneVsOne, roundups, tutorials) to derive from the loop (e.g., filteredPages) and reuse the existing render component/markup to avoid duplication..agents/commands/refresh-compare-pages.md (1)
97-103: Consider requiring source URLs (not just “sources used”) in the command output.This makes refresh PRs auditable and easier to re-verify later.
As per coding guidelines, "Keep command definitions in `.agents/commands/` only. `.claude/commands` and `.cursor/commands` should be symlinks to `../.agents/commands`".Suggested wording tweak
Return a concise summary with: - pages added - pages updated -- sources used +- sources used (include direct URLs and access dates) - validation results - PR URL, or the proposed PR title/body if a PR was not opened🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.agents/commands/refresh-compare-pages.md around lines 97 - 103, Update the refresh-compare-pages command output to require explicit source URLs rather than just "sources used": when producing the summary fields (pages added, pages updated, sources used, validation results, PR URL/proposed PR title/body) include a "sources used" list that contains both source names and their canonical URLs; locate the command definition for "refresh-compare-pages" in .agents/commands/refresh-compare-pages.md and modify the output spec accordingly. Also ensure command placement follows the guideline: keep the authoritative command definition in .agents/commands and replace any definitions in .claude/commands or .cursor/commands with symlinks pointing to ../.agents/commands so those directories only contain symlinks to the single source of truth.apps/marketing/content/compare/superset-vs-conductor.mdx (1)
74-87: Minor formatting inconsistency in list sections.There's an empty line after "Choose Superset if you:" (line 75) but not after "Choose Conductor if you:" (line 81). This creates slight visual inconsistency in the source file, though it may render the same.
📝 Suggested fix for consistency
**Choose Superset if you:** - - Use multiple CLI agents or want the freedom to adopt new ones instantlyOr alternatively, add an empty line after line 81 as well for symmetry.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/content/compare/superset-vs-conductor.mdx` around lines 74 - 87, The two section headers "Choose Superset if you:" and "Choose Conductor if you:" are inconsistently separated from their bullet lists; make them symmetrical by adding a single blank line after the "Choose Conductor if you:" header (the block starting with that header and the bullets underneath) so both headers have one empty line before their lists, leaving the "Verdict:" block unchanged.apps/marketing/content/compare/best-ai-coding-agents-2026.mdx (2)
142-144: Minor grammar: compound adjectives need hyphens.When compound adjectives precede a noun, they should be hyphenated for clarity.
📝 Suggested fix
-**Best for:** Smaller scoped changes, pair-programming style loops, lightweight terminal usage +**Best for:** Smaller-scoped changes, pair-programming-style loops, lightweight terminal usage🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/content/compare/best-ai-coding-agents-2026.mdx` around lines 142 - 144, Update the "**Best for:**" line to hyphenate compound adjectives: change "Smaller scoped changes" to "Smaller-scoped changes" and "pair-programming style loops" to "pair-programming-style loops" (leave "lightweight terminal usage" unchanged). Edit the string containing "**Best for:** Smaller scoped changes, pair-programming style loops, lightweight terminal usage" accordingly.
88-90: Missing horizontal rule separator before section.Other major sections (e.g., lines 40, 72, 184) are preceded by
---separators, but "Detailed Breakdown" is missing one after the comparison table.📝 Suggested fix
| **Conductor** | Orchestrator | Multiple agents, Git worktrees | macOS app | Local Mac app workflow | Public pricing not emphasized on current site | + --- ## Detailed Breakdown🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/content/compare/best-ai-coding-agents-2026.mdx` around lines 88 - 90, Add the missing horizontal rule separator '---' immediately before the "## Detailed Breakdown" heading to match the pattern used for other major sections; locate the "## Detailed Breakdown" heading in the document and insert a line with three dashes directly above it (same style used at other section breaks) so the comparison table and this section are separated consistently.apps/marketing/content/compare/superset-vs-claude-code.mdx (1)
44-44: Minor grammar: "full time" should be hyphenated.When used as a compound modifier, "full-time" should be hyphenated.
📝 Suggested fix
-Claude Code is Anthropic's official coding agent. The CLI is still the core experience, but Anthropic now also documents a Desktop app for macOS and Windows if you want Claude Code without living in the terminal full time. Underneath, it is still the same product: you describe what you want, Claude reads your codebase, writes code, runs commands, and iterates through the task using Anthropic models. +Claude Code is Anthropic's official coding agent. The CLI is still the core experience, but Anthropic now also documents a Desktop app for macOS and Windows if you want Claude Code without living in the terminal full-time. Underneath, it is still the same product: you describe what you want, Claude reads your codebase, writes code, runs commands, and iterates through the task using Anthropic models.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/content/compare/superset-vs-claude-code.mdx` at line 44, The phrase "full time" in the sentence describing Claude Code should be hyphenated as "full-time" when used as a compound modifier; update the text in the paragraph that begins "Claude Code is Anthropic's official coding agent..." (the sentence mentioning "Claude Code without living in the terminal full time") to use "full-time" instead of "full time".apps/marketing/src/lib/compare-utils.ts (1)
44-51: Minor:#stripping may affect legitimate content.The regex
/[*_~>#]/gon line 48 removes all#characters, which could affect answers containing literal#symbols (e.g., "C#", "issue#123", or hex colors like "#fff"). For FAQ content this is likely rare, but consider targeting only leading#for headings:🔧 Optional fix to preserve non-heading `#` characters
function stripMarkdownFormatting(value: string): string { return value .replace(/\[([^\]]+)\]\([^)]+\)/g, "$1") .replace(/`([^`]+)`/g, "$1") - .replace(/[*_~>#]/g, "") + .replace(/^#+\s*/gm, "") + .replace(/[*_~>]/g, "") .replace(/\s+/g, " ") .trim(); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@apps/marketing/src/lib/compare-utils.ts` around lines 44 - 51, stripMarkdownFormatting currently strips all '#' characters via .replace(/[*_~>#]/g, ...) which removes legitimate '#' like "C#" or "#fff"; update the implementation to only remove leading heading hashes instead: replace the global removal that includes '#' with two targeted replacements — one that strips markdown heading markers at line starts (e.g., remove leading 1–6 '#' plus following space using a multiline regex) and a second that still removes other markdown punctuation like '*' '_' '~' and '>' without touching '#' elsewhere; keep the existing link and code-backtick replacements and the whitespace/trim logic intact.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In @.agents/commands/refresh-compare-pages.md:
- Around line 97-103: Update the refresh-compare-pages command output to require
explicit source URLs rather than just "sources used": when producing the summary
fields (pages added, pages updated, sources used, validation results, PR
URL/proposed PR title/body) include a "sources used" list that contains both
source names and their canonical URLs; locate the command definition for
"refresh-compare-pages" in .agents/commands/refresh-compare-pages.md and modify
the output spec accordingly. Also ensure command placement follows the
guideline: keep the authoritative command definition in .agents/commands and
replace any definitions in .claude/commands or .cursor/commands with symlinks
pointing to ../.agents/commands so those directories only contain symlinks to
the single source of truth.
In `@apps/marketing/content/compare/best-ai-coding-agents-2026.mdx`:
- Around line 142-144: Update the "**Best for:**" line to hyphenate compound
adjectives: change "Smaller scoped changes" to "Smaller-scoped changes" and
"pair-programming style loops" to "pair-programming-style loops" (leave
"lightweight terminal usage" unchanged). Edit the string containing "**Best
for:** Smaller scoped changes, pair-programming style loops, lightweight
terminal usage" accordingly.
- Around line 88-90: Add the missing horizontal rule separator '---' immediately
before the "## Detailed Breakdown" heading to match the pattern used for other
major sections; locate the "## Detailed Breakdown" heading in the document and
insert a line with three dashes directly above it (same style used at other
section breaks) so the comparison table and this section are separated
consistently.
In `@apps/marketing/content/compare/superset-vs-claude-code.mdx`:
- Line 44: The phrase "full time" in the sentence describing Claude Code should
be hyphenated as "full-time" when used as a compound modifier; update the text
in the paragraph that begins "Claude Code is Anthropic's official coding
agent..." (the sentence mentioning "Claude Code without living in the terminal
full time") to use "full-time" instead of "full time".
In `@apps/marketing/content/compare/superset-vs-conductor.mdx`:
- Around line 74-87: The two section headers "Choose Superset if you:" and
"Choose Conductor if you:" are inconsistently separated from their bullet lists;
make them symmetrical by adding a single blank line after the "Choose Conductor
if you:" header (the block starting with that header and the bullets underneath)
so both headers have one empty line before their lists, leaving the "Verdict:"
block unchanged.
In `@apps/marketing/src/app/compare/page.tsx`:
- Around line 33-35: The code repeats pages.filter calls and section JSX for
each type (oneVsOne, roundups, tutorials); replace these with a small config
array like [{ key: "1v1", title: "1v1", id: "one-v-one" }, ...] and map over it
to compute filteredPages = pages.filter(p => p.type === cfg.key) and render each
section (use the same rendering logic currently in the page sections around the
existing oneVsOne/roundups/tutorials blocks). Update any local identifiers
(oneVsOne, roundups, tutorials) to derive from the loop (e.g., filteredPages)
and reuse the existing render component/markup to avoid duplication.
In `@apps/marketing/src/components/JsonLd/JsonLd.tsx`:
- Around line 140-142: Normalize the keywords array before injecting into the
JSON-LD object: transform the existing keywords variable by mapping each entry
to its trimmed string and filtering out empty/falsy results (e.g., keywords =
keywords.map(k => k.trim()).filter(Boolean)), then only spread into the payload
using the existing conditional spread (the place that currently does
...(keywords && keywords.length > 0 && { keywords })); update the code around
the JsonLd.tsx component where keywords is used so the JSON-LD contains only
non-empty, trimmed keyword values.
In `@apps/marketing/src/lib/compare-utils.ts`:
- Around line 44-51: stripMarkdownFormatting currently strips all '#' characters
via .replace(/[*_~>#]/g, ...) which removes legitimate '#' like "C#" or "#fff";
update the implementation to only remove leading heading hashes instead: replace
the global removal that includes '#' with two targeted replacements — one that
strips markdown heading markers at line starts (e.g., remove leading 1–6 '#'
plus following space using a multiline regex) and a second that still removes
other markdown punctuation like '*' '_' '~' and '>' without touching '#'
elsewhere; keep the existing link and code-backtick replacements and the
whitespace/trim logic intact.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 6920aeff-7df1-4dc7-92cf-96fbf45b68cd
📒 Files selected for processing (20)
.agents/commands/refresh-compare-pages.mdapps/marketing/content/compare/best-ai-coding-agents-2026.mdxapps/marketing/content/compare/best-terminal-for-ai-coding.mdxapps/marketing/content/compare/multiple-claude-code-agents-parallel.mdxapps/marketing/content/compare/superset-vs-claude-code.mdxapps/marketing/content/compare/superset-vs-codex.mdxapps/marketing/content/compare/superset-vs-conductor.mdxapps/marketing/content/compare/superset-vs-cursor.mdxapps/marketing/content/compare/superset-vs-devin.mdxapps/marketing/content/compare/superset-vs-github-copilot.mdxapps/marketing/content/compare/superset-vs-opencode.mdxapps/marketing/content/compare/superset-vs-t3-chat.mdxapps/marketing/content/compare/superset-vs-warp.mdxapps/marketing/content/compare/superset-vs-windsurf.mdxapps/marketing/docs/SEO_COMPARISON_PAGES_PLAN.mdapps/marketing/src/app/compare/[slug]/components/CompareLayout/CompareLayout.tsxapps/marketing/src/app/compare/[slug]/page.tsxapps/marketing/src/app/compare/page.tsxapps/marketing/src/components/JsonLd/JsonLd.tsxapps/marketing/src/lib/compare-utils.ts
💤 Files with no reviewable changes (1)
- apps/marketing/docs/SEO_COMPARISON_PAGES_PLAN.md
Kitenite
changed the title
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.
Summary
best-terminal-for-ai-coding,multiple-claude-code-agents-parallel, andsuperset-vs-t3-chat, plus the compare-page route/index metadata needed to surface roundups and tutorials correctly.agents/commands/refresh-compare-pages.mdworkflow for fresh-research compare updatesWhy / Context
The compare pages were starting to undersell the current Superset product by describing it too narrowly as only a terminal/orchestrator.
This PR shifts the compare content to the actual shipped product story and adds a repeatable update workflow for future refreshes.
How It Works
The compare route now understands page type distinctions beyond plain 1v1 comparisons, so roundups/tutorials can be listed and labeled correctly. The slug page also emits richer structured metadata (FAQ and breadcrumb JSON-LD plus page keywords from MDX content).
On the content side, the refreshed compare pages consistently position Superset as a local-first agent workspace with worktrees, chat, diff/file review, browser preview, and MCP tooling, while keeping cloud language cautious and limited to publicly supportable claims. The new refresh command codifies the expected workflow: use fresh official sources, update stale pages in the same run, and make the PR explicit about what was added vs refreshed.
Manual QA Checklist
Testing
bun run --cwd apps/marketing typecheckbun -e 'import { getComparisonPages } from "./src/lib/compare"; const pages = await getComparisonPages(); console.log(JSON.stringify({count: pages.length, slugs: pages.map((p)=>p.slug).sort()}, null, 2));'(fromapps/marketing)Known Limitations
bun run --cwd apps/marketing buildafter the final cleanup. Earlier in this workstream, the marketing build compiled app code but failed during unrelated page-data collection because the repo environment is missing the Neon connection string needed by non-compare routes like/_not-foundand/terms.