code-auditing

Audit code quality, security, and ecommerce domain rules with scoped evidence-based reviews.

Updated Jun 11, 2026
One-click install
npx skills add https://github.com/Marcel-Carrillo/ProyectosIA --skill code-auditing-marcel-carrillo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: code-auditing
Source: https://github.com/Marcel-Carrillo/ProyectosIA/tree/main/ai-specs/skills/code-auditing
Command: npx skills add https://github.com/Marcel-Carrillo/ProyectosIA --skill code-auditing-marcel-carrillo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Manual code reviews often miss security vulnerabilities, dead code, and domain rule violations, especially in ecommerce systems where supplier data exposure or status mixing can cause serious business issues. This Skill provides a systematic, evidence-based audit workflow that scopes reviews to the task at hand instead of scanning the entire repository. ## Core Features & Use Cases - Three Audit Modes: Run task-focused audits on a single change or pull request, targeted audits on a specific area (backend, frontend, security, data model), or comprehensive repository-wide audits only when explicitly requested. - Ecommerce Domain Checks: Verify CustomerOrder vs SupplierOrder separation, supplier data exposure prevention, ProductVariant sellable-unit rules, and status isolation across payment, fulfillment, shipment, return, and refund flows. - Dead Code and Security Analysis: Detect unused imports, exports, and dependencies with tools like knip, plus check for hardcoded secrets, injection risks, and missing authorization. - Use Case: After implementing a new supplier order endpoint, run a task-focused audit to confirm supplier costs are not exposed in customer-facing API responses and that tests cover validation and authorization cases. ## Quick Start Audit the changes in my current git diff for security issues and ecommerce rule violations, then generate a prioritized findings report.

Frequently Asked Questions about code-auditing

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a code audit on a single pull request or change?▼

Use the task-focused audit mode, which is the default. It inspects only changed files, directly related supporting files, relevant tests, and OpenSpec artifacts, avoiding a full repository scan.

What tools detect dead code in JavaScript and Python projects?▼

Knip detects unused exports, files, and dependencies in JavaScript and TypeScript via npx knip, while the deadcode package handles Python. All tool findings must be manually verified to filter false positives like dynamic imports and framework conventions.

Does this audit require GitHub, Jira, or Sentry integrations?▼

No, the audit works fully offline using local files, git diff, test output, and documentation. External tools like GitHub, Context7, Jira, or Sentry are optional and used only when explicitly relevant to the task.

When should I run a comprehensive repository-wide audit?▼

Only when the user explicitly requests a full audit. Comprehensive audits consume significant time and context, so task-focused or targeted area audits are preferred for normal development workflows.

What ecommerce-specific issues does the audit check for?▼

It verifies supplier costs and credentials are never exposed through customer-facing APIs, CustomerOrder and SupplierOrder concepts stay separate, ProductVariant is the sellable unit, and payment, fulfillment, shipment, return, and refund statuses are not mixed.