Engineering Standards

Inspection Methodology & Accuracy Principles

At B A C Web Development Services, we believe intelligence tools must be grounded in empirical evidence, cryptographic rigor, and strict boundary ethics. Here is how our engine operates.

1. Evidence-Based Technology Identification

We never fabricate results or assume an entire architecture based on superficial hints. Every detected technology must satisfy explicit rules against retrieved artifacts:

  • High Confidence: Explicit declaration via HTTP response headers (e.g., Server: nginx, X-Powered-By: Next.js), formal <meta name="generator"> tags, or definitive script bundle paths.
  • Medium Confidence: DOM structure patterns, characteristic CSS class naming conventions, or distinctive inline JSON variables.
  • Low Confidence: Generic circumstantial markers or third-party tracking scripts that frequently appear across diverse platforms.

2. Server-Side Request Forgery (SSRF) Guardrails

Scanning arbitrary user-provided URLs introduces severe security risks to internal networks. B A C Tech Intelligence enforces defense-in-depth protections on every outbound request:

  • Only standard HTTP (80) and HTTPS (443) protocols are permitted; arbitrary ports and raw schemes are immediately rejected.
  • DNS pre-resolution inspects both IPv4 and IPv6 results. Any destination falling into loopback (127.0.0.0/8), private (RFC 1918), link-local, carrier-grade NAT, or cloud metadata ranges (169.254.169.254) is blocked.
  • Target IPs are pinned during connection to defeat DNS rebinding attacks.
  • Redirects are traversed manually hop-by-hop, with each destination re-validated against the SSRF guardrail before connection.
  • Downloaded content is capped at 5MB with strict decompression limits to neutralize decompression bombs.

3. Observations vs. Vulnerability Claims

A missing response header (such as Content-Security-Policy or Permissions-Policy) is a defense-in-depth observation, not proof of an exploitable vulnerability. We provide objective posture grades without alarmist rhetoric or unsupported severity claims.

4. Persistence and Database Inference

Backend persistence layers (such as PostgreSQL, MySQL, Redis, or MongoDB) reside securely behind internal firewalls and are rarely observable over public HTTP. B A C Tech Intelligence explicitly marks database components as "Unconfirmed / Private" unless indisputable public evidence is exposed.