PhpCodeArcheology

Why STEADY

STEADY (77) because PhpCodeArcheology is a thorough, transparent architecture analyzer (60+ metrics with a formula glossary, 14 detectors, six formats incl. SARIF), is framework-aware to avoid false positives, and is genuinely agent-native — a native MCP server plus a generated CLAUDE.md. Held at 77 because it is PHP-only by design, the headline counts and 'first-with-MCP' claim are self-reported rather than benchmarked here, and its thresholds are self-authored.

Public-surface checklist

• PASS Homepage loads (required)
• PASS Primary value prop (required) — PHP architecture metrics + problem detectors, with native MCP querying
• PASS Cta present (required) — Get Started / composer require / View on GitHub
• PASS Pricing or access — MIT license; composer require --dev; no extensions required
• PASS Evidence or demo — Feature breakdown, metric glossary, report-format and MCP-tool examples

What we saw

4 screenshots captured by the Hlido engine during the reviewed run (run-2ea5b5fac0ede252-phpcodearcheology-github-io). Our own captures — not vendor marketing material.

×

←→

What it does well

• 60+ architecture/maintainability metrics per file/class/method/function, with a glossary of formulas and thresholds
• 14 problem detectors (god classes, dead code, circular dependencies, deep inheritance) with severity and actionable descriptions
• Native MCP server — query hotspots, dependency graphs and refactoring priorities in natural language
• Generates a CLAUDE.md project overview for an AI coding assistant to read first
• Six output formats including SARIF for GitHub Code Scanning and JSON for CI
• Framework-aware: reads composer.json (Symfony/Laravel/Doctrine) and adjusts thresholds to cut false positives
• MIT, PHP 8.2–8.5, composer-installed, no extensions required; test-coverage integration (PHPUnit/Pest/Codeception)

What it fails at

• PHP-only by design — no value for other languages
• Headline figures (60+ metrics, 14 detectors, 'first PHP tool with native MCP') are project claims, not benchmarked here
• Metric thresholds are self-authored, as with any single-vendor analyzer
• Depth on very large legacy codebases is asserted rather than demonstrated on the reviewed surface

Pricing & access

• Pricing findable on the public surfacePASS MIT license; composer require --dev; no extensions required (tested 2026-08-26)

Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-08-26.

Compared to

• Sonarly php-architecture-depth-vs-general-quality

  Sonarly is a general code-quality/analysis play; PhpCodeArcheology is PHP-specific, architecture-focused (LCOM, coupling, instability) and ships a native MCP server plus a generated CLAUDE.md. Sonarly for broad quality, PhpCodeArcheology for deep PHP architecture with agent-native querying.

• Serena metrics-and-health-vs-structural-navigation

  Serena is a multi-language semantic code toolkit for navigation/editing; PhpCodeArcheology measures PHP architecture health and exposes it over MCP. Serena for cross-language structural operations, PhpCodeArcheology for PHP metrics and refactoring priorities.

Agent relevance

CLI MCP Behavioral-testable

Agentic-Commerce Readiness 65/100 · INTEGRABLE

Independent readiness for agent delegation & transaction. How it’s scored · check live

Two agent paths: a native MCP server the assistant queries in natural language for hotspots, dependency graphs, health scores and refactoring priorities (MCP tool calls such as get_hotspots and get_refactoring priorities), and a generated CLAUDE.md the assistant reads before editing the codebase. Also runs as a composer-installed CLI producing JSON/SARIF for CI. Directly designed for AI-assisted PHP maintenance.

Agent-friendly score: 8/10

Evidence

• 60+ metrics and 14 problem detectors across files, classes and methods — source (2026-08-26) verified
• Ships a native MCP server and generates a CLAUDE.md for AI assistants — source (2026-08-26) verified
• Six report formats including SARIF and JSON; framework-aware via composer.json; MIT, PHP 8.2–8.5 — source (2026-08-26) verified