Compileq
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Compileqscan my repo for GDPR compliance risks"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Compileq — AI Compliance Scanner and MCP Server
Built by Psylinks Security Private Limited
Compileq scans a codebase for likely compliance gaps — GDPR, CCPA/CPRA, WCAG, PCI-DSS, HIPAA, EU AI Act, LGPD, COPPA, PIPEDA, and more via an official-source registry covering 24+ regulations worldwide — scores the risk, estimates fine exposure, and proposes concrete fixes. It runs as a Claude Skill or as an MCP server, so it works inside Claude Code, Claude Cowork, Cursor, VS Code + Copilot, Windsurf, or any other MCP-compatible IDE.
⚠️ This repository is public for product discovery, but the software is proprietary, not open source. Public visibility does not grant permission to copy, fork, redistribute, sublicense, or create derivative works. See
LICENSE. Copyright © Psylinks Security Private Limited. All rights reserved. Do not fork, redistribute, or publish this code publicly. SeeLICENSEfor the full terms. If you found this repository without authorization, please contact us at psylinkssecurity.com.
What it does
Runs a static, rule-based scan of your repository for patterns that commonly correlate with compliance gaps (missing consent gates before tracking scripts, unencrypted PII/PHI fields, raw card numbers touching your own storage, missing alt text and form labels, automated decision-making without human oversight, and more).
Scores every finding: a risk score (0–100), a confidence score (0–1), and a rough fine-exposure estimate — see
docs/RISK_SCORING.mdfor the exact formulas.Produces a category readiness score per regulation and an overall readiness score, in both Markdown (for humans) and JSON (for tooling).
When it doesn't know something, it doesn't guess: it routes through an official-source registry (
rules/official_sources.yaml) that maps each regulation to its actual government/regulator/standards-body domain, and explicitly refuses to treat law-firm blogs, marketing pages, or aggregators as authoritative — seedocs/ARCHITECTURE.md.This is a triage tool, not legal advice or a certification. Every report says so, and confidence scores exist precisely so low-certainty findings are never presented as fact.
Related MCP server: mcp-security-compliance
Why teams use Compileq
Compileq is designed for founders, engineering teams, security reviewers, privacy teams, accessibility specialists, and AI-assisted development workflows that need an early compliance signal before launch. It helps turn questions such as “are we GDPR-ready?”, “does this signup flow create CCPA risk?”, “are our payment forms PCI-DSS-sensitive?”, and “does this AI feature need EU AI Act review?” into a traceable engineering triage report.
It is useful for privacy compliance scanning, GDPR and CCPA readiness reviews, WCAG accessibility checks, PCI-DSS payment-flow triage, HIPAA-sensitive code review, EU AI Act preparation, LGPD/COPPA/PIPEDA checks, and MCP-enabled AI code review in Claude, Cursor, Codex, VS Code, Copilot, Windsurf, and other compatible tools.
Quick start
pip install -r scanner/requirements.txt
python -m scanner.cli scan /path/to/your/repo --markets EU,US-CA --out report.mdAs a Claude Skill (Claude Code / Claude Cowork)
cp -r compileq ~/.claude/skills/compileq
pip install -r ~/.claude/skills/compileq/scanner/requirements.txtClaude will pick up SKILL.md automatically — just ask it to scan a repo
for compliance issues.
Codex, Cursor, Antigravity, and other AI IDEs
For OpenAI Codex or any Agent Skills-compatible tool, copy the package into its documented skill directory. For example:
mkdir -p ~/.codex/skills/compileq
cp -R . ~/.codex/skills/compileq/Cursor can discover the SKILL.md package through a compatible skills folder,
or use the included .cursor/rules/compileq.mdc
project-rule bridge. For Antigravity and tools that do not auto-discover
Agent Skills, copy AGENTS.md into the project root or paste
its instructions into the IDE's project rules. The scanner and MCP server work
independently of the host AI tool.
As an MCP server (Cursor, VS Code + Copilot, Windsurf, Claude Desktop)
pip install -r mcp_server/requirements.txtAdd to your MCP host's config:
{
"mcpServers": {
"compileq": {
"command": "python",
"args": ["/absolute/path/to/compileq/mcp_server/server.py"]
}
}
}Full install instructions: docs/INSTALL.md.
Coverage
Regulations with active, scored pattern rules | GDPR, CCPA/CPRA, WCAG, PCI-DSS, HIPAA, EU AI Act, LGPD, COPPA, PIPEDA |
Regulations in the official-source registry | 24+, spanning EU/UK/US/Brazil/Canada/South Africa/Singapore/India/China/Japan/South Korea/Australia/Nigeria plus global standards (SOC 2, ISO 27001) |
Full matrix: docs/COVERAGE.md.
Repository layout
SKILL.md Claude Skill definition
scanner/ the engine (offline, deterministic, testable)
engine.py repo walker + rule matcher
scoring.py risk / confidence / fine-exposure formulas
report.py Markdown report generator
web_lookup.py official-source allowlist enforcement
fallback.py out-of-knowledge lookup protocol
branding.py product/maker attribution (single source of truth)
cli.py command-line entry point
rules/*.yaml pattern rules, one file per regulation
rules/official_sources.yaml the 24-regulation official-domain registry
mcp_server/server.py MCP server wrapper for non-Claude IDEs
docs/ architecture, scoring formulas, coverage, install
tests/ 27 tests, all fixtures includedDocumentation
docs/ARCHITECTURE.md— how the pieces fit together, including the out-of-knowledge protocoldocs/RISK_SCORING.md— every formula behind every number in a reportdocs/RULES_SCHEMA.md— how to add a new rule or regulationdocs/COVERAGE.md— active vs. registry-only regulation coveragedocs/INSTALL.md— detailed install steps for every IDEdocs/ROADMAP.md— what's nextCONTRIBUTING.md— internal contribution guide (authorized contributors only)
Testing
python -m unittest discover -s tests -v27 tests covering the rule loader, scoring math, engine detection against real fixture files, and the official-source allowlist logic.
License
Proprietary — All rights reserved, Psylinks Security Private Limited. See
LICENSE. This software is not licensed under MIT, Apache,
or any open-source license; redistribution and derivative works are
prohibited without written permission.
Compileq is a product of Psylinks Security Private Limited — psylinkssecurity.com
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Zero-config MCP security scanner for AI-generated apps. 25K+ vulnerability patterns.
EU compliance corpus across 8 frameworks (NIS2, DORA, AI Act, ISO 27001 + more) via MCP.
Compliance frameworks (SOC 2, ISO 27001, CMMC, NIST, more) delivered to AI agents as MCP tools.
AI legal compliance: contract review, risk scoring, EU/CN AI act, watermark check. 8 MCP tools.
Related MCP Servers
- AlicenseBqualityCmaintenanceBrings GDPR compliance knowledge directly into your IDE, enabling developers to identify and address data protection requirements early in the development lifecycle.342MIT
- AlicenseNot gradedqualityBmaintenanceIntegrates authoritative security compliance frameworks (ISO 27001, NIST 800-53, OWASP ASVS, NIST SSDF) into AI-assisted development, offering control lookups, cross-framework mappings, build-time guardrails, and automated audit evidence generation.1743MIT
- AlicenseAqualityDmaintenanceAn MCP server for web accessibility testing that enables scanning, auditing, and fixing WCAG, ADA, and other compliance issues directly from your IDE, with free local scans, AI-generated framework-aware fixes, and verification capabilities.14118MIT
- AlicenseNot gradedqualityAmaintenanceEnables AI coding tools to scan projects for security vulnerabilities, hardcoded secrets, injection flaws, and privacy violations with 699 rules and 76 MCP tools, all running locally with zero telemetry.216MIT