Skip to content
EDITORIAL STANDARDS

Editorial Policy

WebPulse is an intelligence platform, not an opinion blog. Every claim is backed by data, every source is cited, and every process is documented here.

Our editorial mission

WebPulse publishes data-driven analysis about the health, security, and future-readiness of web frameworks. Our audience is technology decision-makers: CTOs, engineering leaders, and the executives who sign infrastructure budgets. We write for clarity and accuracy, not for clicks or controversy.

Data sources and verification

All quantitative claims on WebPulse are derived from verifiable public sources. We do not rely on surveys, vendor-supplied statistics, or unverifiable third-party reports.

Primary data sources
  • NIST National Vulnerability Database (NVD) — CVE counts, CVSS severity scores, and vulnerability timelines. Cross-referenced with the CISA Known Exploited Vulnerabilities catalog.
  • GitHub REST API — Repository stars, contributor counts, release cadence, issue resolution times, and pull request merge velocity. Collected programmatically on a daily schedule.
  • Common Crawl WARC archives — Large-scale framework detection across publicly accessible web pages.
  • Tranco top-100K list — Traffic-ranked domain scanning for framework and generation classification.
  • Live site scans — Real-time framework detection submitted by users through our scanning tools.
Verification process
  • Raw data is collected via automated scripts and stored in version-controlled JSON files.
  • Scores are computed algorithmically from raw data using a documented, weighted methodology (see Methodology).
  • Statistical claims in stories (e.g., CVE counts, adoption percentages) are validated against source data before publication.
  • When we cite a number, the underlying data can be traced to the specific API call or dataset that produced it.

Editorial standards

Grounded in present data

We report on what the data shows today. We do not publish speculative predictions, forecasts, or "what might happen" content. When we describe trends, we cite the underlying data points.

No vendor influence

No framework vendor, hosting provider, or technology company pays for scores, editorial coverage, or placement on WebPulse. Our revenue comes from Adyog's security business, not from the platforms we analyze.

Statistical honesty

We do not cherry-pick data to support a predetermined conclusion. When a framework scores well, we report it. When a widely used framework has serious vulnerabilities, we report that too. Context and nuance matter — a high CVE count for a large ecosystem means something different than a high CVE count for a niche tool.

Clear scope

WebPulse analyzes web frameworks, not the companies or individuals who build them. Our scoring reflects the technical characteristics of the software, not opinions about the people behind it.

Audience-appropriate language

We write for decision-makers, not developers. Technical findings are translated into business impact. Jargon is minimized. When technical terms are necessary, they are explained.

How we use AI

Many WebPulse news articles are written by an automated newsroom that uses AI. They carry the byline WebPulse Newsroom and an "AI-assisted" label. Articles written or edited by a person carry that person's name.

How a newsroom article is made. Our system reads a wide set of security, AI and web-platform sources, picks stories relevant to technology decision-makers, and fetches the full original article. Claude, an AI model from Anthropic, writes the story from that source and, where relevant, our own scan data, and must attribute every fact it uses.
How it is checked. A separate AI review checks each draft against the same source: claims the source does not support are treated as errors, and misleading framing, overstatement and pure restatement of the source block publication. Automated checks also reject drafts that reuse the source's wording, fall below minimum length and sourcing, repeat a story we have already covered, or break our language rules. A draft that fails is revised or discarded.
What is published automatically. Drafts that pass every check are published without a person reading them first. Each one credits and links the original reporting it is based on.
What AI does not do. AI does not compute our framework scores or census figures; those come from our scanners and public data such as NIST NVD and CISA.
Images. Many article images are generated with AI to illustrate the story's idea, and each one is labelled "AI-generated image" beneath it. Other images are licensed photographs credited to their photographer. We never present an AI-generated image as a real photograph of an event.
Chat assistant. The chat assistant on this site uses Google Gemini to answer questions about our data.
Corrections. If you find an error in any article, tell us through the contact details below and we will correct it and note the change.

Corrections policy

We take factual accuracy seriously. If we publish an error, we correct it promptly and transparently.

  • Minor errors (typos, broken links, formatting) are corrected without notice.
  • Factual errors (incorrect statistics, misattributed data, misleading framing) are corrected with a note at the top of the affected content indicating what changed and when.
  • Significant retractions result in the story being updated with a prominent correction notice, or removed if the core claim is unfounded.

To report an error or request a correction, email [email protected] with the URL of the affected page and a description of the issue.

Editorial independence

WebPulse is operated by Adyog, a security company. The platform is funded by Adyog's commercial security business, not by advertising, sponsorships, or vendor payments.

This structure is intentional: because WebPulse does not depend on the frameworks it covers for revenue, there is no financial incentive to favor or suppress any particular technology.

Adyog's own technology choices (Python, FastAPI, HTMX) are scored by the same methodology as every other framework. We do not give ourselves preferential treatment.

Content types

WebPulse publishes several types of content, each held to the same factual standards:

Data stories Analysis built around specific data findings. Every claim cites the underlying dataset.
Industry reports Sector-specific analysis (healthcare, finance, government, etc.) combining scan data with regulatory context.
Regional editions Country and region-specific coverage with localized scan data and regulatory analysis.
Framework scores Algorithmically computed scores with no editorial discretion in the calculation.

Questions about our editorial process

If you have questions about how a specific story was produced, how a score was calculated, or how our data was collected, contact us at [email protected]. We are happy to explain our work.