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 We do not publish speculative predictions, forecasts, or "what might happen" content.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.

AI-assisted content disclosure

WebPulse uses AI tools in its editorial workflow. We believe transparency about AI use is essential for reader trust.

What AI does

Draft generation: Some stories begin as AI-generated drafts based on collected data. The AI (Google Gemini) synthesizes data points into narrative form.
Adversarial review: AI is used as a fact-checking layer, challenging claims and flagging unsupported assertions before publication.
Data summarization: Large datasets are summarized with AI assistance for regional and industry analysis pages.
Chat assistant: The on-site chat uses Google Gemini to answer visitor questions about web framework data.

What AI does not do

AI does not determine scores. Scores are computed algorithmically from source data.
AI does not publish content autonomously. All content passes through editorial guardrails and confidence thresholds before publication.
AI-generated claims are validated against the underlying data. If a claim cannot be traced to a verifiable source, it is removed.

Quality control

Auto-generated stories must pass editorial guardrails (hard rules on sourcing, tone, and accuracy) and an AI adversarial review before publication. Stories that fail either check are held for manual revision or killed. The system is designed to be conservative — we would rather miss a story than publish a bad one.

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.