Kapitol Review: strong product, fixable gaps (80/100)

Kapitol.ai is a freemium web and API product for curated U.S. congressional stock-trade intelligence and alerts.

Fact	Detail
Domain	kapitol.ai
Category	Curated U.S. congressional stock-trade intelligence and alerts
Pricing	Freemium; $16.67–$35 USD range; Premium at $35/month, $300 yearly, or $600 every three years; seven-day trial for $1
Pages crawled	40
Crawl date	2026-10-21

Kapitol.ai earns 80/100 because it turns public congressional-trade disclosures into a clearer, interpreted service for self-directed investors and market watchers. The strongest evidence is its Usability score of 88.9/100 and Docs & self-serve help score of 88.6/100: the crawl found a REST reference, a historical-data guide, and practical Python guides, while the product pages explain source data, alert timing, price, cancellation, and the free-database alternative. Technical SEO (86.9/100), Risk & stability (86.3/100), and Decision-support surfaces (84/100) reinforce the strength of the public experience.

The result is not uniformly polished. Design execution scored 69/100, Editorial QA of content 68.8/100, Accessibility 74/100, and Performance 74.9/100. The clearest measured issue is the mobile API documentation route’s 5.6 s LCP, above the 2.5 s target. Kapitol’s next gains are practical: improve that route, tighten visual and content QA, and preserve the strong explanatory work.

Kapitol.ai is immediately recognizable as a congressional-trade intelligence service, but its first screen names the subject more clearly than the buyer. The homepage says "Track every trade Congress makes. Know which ones matter," then explains that trades are flagged when they align with committee work. The comparison page adds the sharper audience definition: investors seeking a smaller curated signal feed with scores, context, tracking, and email delivery. Quantitative proof is substantial—740 trades filed in the last 30 days, 105K+ disclosures, 535 members, and 5,000+ companies—but the +24.4% average return per Signal versus +14.6% for the S&P 500 needs adjacent methodology. Put "curated, scored congressional-trade signals for investors" in the hero and link the methodology beside the return claim.

Kapitol.ai gives prospects unusually usable answers about the product, price, source data, alert timing, cancellation, and the free-database alternative. Pricing is concrete: a 7-day trial for $1, followed by $35 monthly, $300 yearly, or $600 every three years. The app FAQ says data comes from official House Clerk and Senate filing systems, alerts usually arrive within minutes of public filing, and the product is information rather than investment advice. The remaining decision gap is whether the curation premium is justified. The pricing page presents +24.4% average return per Signal against +14.6% for the S&P 500, but the supplied evidence does not show the sample, dates, fees, or benchmark method beside it. Add that explanation and state the investor audience above the fold.

Kapitol.ai communicates its core offer quickly and gives the main actions plain-language labels. The homepage leads with congressional-trade tracking and a "See Congress Trades" path; pricing offers "Get Started" and a 7-day trial for $1; the contact form uses labeled Name, Email address, and Message fields; and the developer route exposes "Start free" and "Read the docs." The friction is findability for secondary goals. The primary header offers "See Trades" while Pricing appears in the footer inventory, and the supplied mobile homepage places Pricing, Contact, and API Docs after substantial scrolling. Add Pricing beside the hero CTA or to primary navigation, then provide compact mobile shortcuts. Preserve the short contact form and current CTA language.

Kapitol.ai has strong structural foundations on the homepage and pricing page: both captured Lighthouse accessibility runs score 100, the homepage exposes header, navigation, main, and footer landmarks, and it includes a skip link with labeled visible fields. The verified gaps are concentrated in image decisions and the API documentation template. The rendered DOM reports 2 images without an alt decision on each of the homepage, contact, pricing, and articles pages, so each needs a purpose-based choice between descriptive alt text and alt="". On /docs/api, the evidence reports no main landmark, an unnamed button, and a role missing required ARIA attributes. Add the landmark, repair the computed button name, and complete or remove the affected role before relying on the strong marketing-page baseline.

Kapitol.ai has a coherent visual foundation: the homepage uses one clear H1, a green primary CTA, dark navy and mint-green surfaces, rounded cards, and a 44px H1 against 16px body text. At 390px, the crawl measured no horizontal overflow, with scrollWidth390 equal to 390. Mobile interaction hygiene is weaker. Forty small tap targets were reported, including a 40x40px menu button and 32px-high header, footer, and legal links; the captured input also uses 14px text instead of the 16px mobile threshold. Mobile pricing recorded CLS 0.104 with two layout shifts, while desktop pricing recorded 0.01. Raise compact controls to at least 44x44px, increase mobile input text to 16px, and reserve space for the shifting elements. Treat the 529 colors and 75 font sizes as audit signals, not proof of visible inconsistency.

Kapitol.ai is quick on the sampled homepage but has a material mobile documentation outlier. PageSpeed Insights measured homepage LCP at 1.7 seconds, while /docs/api reached 5,579.769 ms LCP; the same route carries an estimated 4,190 ms of render-blocking savings and 327 KiB of unused JavaScript. Mobile pricing also recorded CLS 0.104, just above the 0.1 good threshold, while desktop pricing recorded 0.010. These are lab measurements, not real-user p75 values, because no CrUX field data was supplied. Start with the documentation route: defer non-critical resources, keep only initial-view styles in the first response, and split interactive examples from the initial bundle. Then resize politician images served at 400–500px for 38px slots and apply long caching to fingerprinted assets.

Kapitol.ai writes most clearly when it connects promise to mechanism. "Kapitol.ai reads every stock trade members of Congress disclose, flags the ones that line up with their committee work, and explains why" is specific, while /developers adds a 90s poll interval, 2012 as the earliest trade year, named endpoints, amount ranges, and filer comments. The homepage, app, and pricing paragraphs average 9.2, 11.3, and 15.9 words respectively, supporting scanability. Editorial hygiene drops on the edges: the crawler records literal 404 targets polUrl(p.slug) and stockUrl(s.ticker), and /docs/api begins with H4 before H1 and has no recorded meta description. The +24.4% versus +14.6% return comparison also needs cohort, window, weighting, and benchmark context. Fix the URLs, repair the docs shell, and place that method beside the claim.

Kapitol.ai's comparison page gives buyers an honest two-path decision: choose Capitol Trades for breadth and free public browsing, or choose Kapitol.ai for a curated signal feed with scoring, explanations, tracking, and email alerts. Its 12-row table covers price, browsing breadth, history, analysis, and scoring, so the guidance has functional substance. The pricing page is equally explicit about totals—$35 monthly, $300 billed yearly, or $600 every three years—but its "POPULAR" yearly label and 29% saving do not explain which cadence fits a flexible tester, a committed investor, or a long-term buyer. Add one sentence per cadence and a compact "How we compared" note with evidence date, normalized axes, and source boundaries. On mobile, reposition the cookie panel so it does not cover the premium card.

The supplied vulnerability scan found no sitewide robots block, accidental noindex pattern, redirect loop, mass 5xx signal, or raw-versus-rendered content gap. The clearest preventable risk is template leakage: /polUrl(p.slug) and /stockUrl(s.ticker) are literal 404 paths and should be removed from emitted links. Kapitol.ai also has changing-search-result exposure because its finance editorial surface combines research articles with data surfaces; consistent answer-first structure and entity markup are sensible safeguards, not evidence of traffic loss. Historical diagnosis is limited: Wayback supplied only 3 snapshots from 2026, and Google Search Console and SERP data were unavailable. Connect GSC and segment clicks and impressions by page type before attributing future changes to an algorithm update. In the meantime, add a generated-link check and recrawl after removing both placeholders.

Kapitol.ai reads as edited product marketing rather than generic brochure copy. The homepage explains that official disclosures are filtered, scored, and explained, while /developers adds a 90-second poll interval, first_seen_at, named endpoints, amount ranges, and a concrete JSON response. The release risks are observable and fixable. The pricing page states +24.4% average return per Signal versus +14.6% for the S&P 500, and the newsletter says 68% of Signals are in profit, but the supplied copy shows no source or methodology at the point of use. The crawler also confirms one FAQPage block on /app, yet /docs/api has H4 before H1, no recorded canonical, and no meta description; literal polUrl(p.slug) and stockUrl(s.ticker) links return 404. Add claim-level provenance, block template expressions in rendered hrefs, and run heading, metadata, and link checks before release.

Kapitol.ai has a functional developer documentation system rather than a thin sales promise. The crawl found a REST reference with endpoints, schemas, bearer authentication, and a live API base URL, plus a historical-data guide and practical Python material. The Python guide contains 107 code blocks and the historical guide 28; both are marked updated September 10, 2026. The main friction is answer-finding on the developer landing page: it has a sidebar but no search box or breadcrumbs, and its heading structure is a marketing-led overview rather than a task index. Add visible search, breadcrumbs, and a compact route-by-task index, then preserve explicit links from each guide to its endpoint and schema. The published llms.txt summary and TechArticle and FAQPage structured data are useful supporting surfaces as the documentation grows.

Kapitol.ai is crawlable, HTTPS-enabled, and largely server-rendered. robots.txt allows the public site and declares a verified 212-URL sitemap; HTTP and www roots consolidate to HTTPS in one 301 hop; a nonexistent-URL probe returns a real 404; and raw and rendered word counts match across all 40 crawled pages. The main technical defects are template leakage and a documentation performance outlier. /polUrl(p.slug) and /stockUrl(s.ticker) are literal 404 paths, while mobile /docs/api records 5,579.8 ms LCP and 4,190 ms of estimated render-blocking savings. The /articles/ trailing-slash variant also serves a second 200 URL, and HSTS was not observed in sampled HTTPS headers. Remove the placeholders, choose one slash policy, confirm HSTS at the edge, and retest the docs route below 2.5 seconds. Full sitemap coverage remains unverified beyond the 40-page crawl budget.

Kapitol.ai is a strong fit for self-directed investors who want scored, explained congressional-trade signals, and for developers who need a REST API with historical-data guidance. It is less compelling for buyers who only need the broadest free disclosure database; the product itself points those users to Capitol Trades.

The highest-value fixes are clear: bring the mobile API documentation route’s 5.6 s LCP closer to the 2.5 s target, and strengthen the visible content-QA process behind performance comparisons. Design execution (69/100), Editorial QA of content (68.8/100), Accessibility (74/100), and Performance (74.9/100) are the main fair-scoring areas around the strong overall result. Pricing is legible: freemium access, Premium at $35/month, and a seven-day $1 trial.

This is a 13-dimension review based on the 40-page crawl from 2026-10-21, public page content, product facts, and supplied lab measurements. The review uses the public score table; this article frame does not add owner-only findings or re-score dimensions. Sources include Kapitol.ai’s public pages, including the homepage, pricing, app, and developer documentation, plus the supplied crawl and PageSpeed measurements.

Dimension 13, Review-content integrity, was excluded as not applicable. Google Search Console enrichment was not connected or available, so no Search Console findings are included. Read How SiteList scores for the review method.