ai-pro_skill

Overview

This skill is a professional AI and system orchestrator optimized for January 2026 stacks. It combines advanced agent orchestration (GPT-5 and autonomous agents), identity/security with Auth.js v5, payment integration with Stripe v13, and large-codebase intelligence for monorepos and archive search. It targets senior AI engineers who need reliable, secure, and high-performance automation across CI/CD and production systems.

How this skill works

The skill provides patterns, utilities, and protocols to observe system state, reason about errors, check constraints, and act (O.R.C.A.). It inspects API errors, authentication configuration, and codebase context to generate diagnostics, automated retries, or remediation actions. It also supplies high-speed search and context-packing tooling (ripgrep, repomix) to construct compact, relevant context for agent decision-making.

When to use it

• Automating self-healing integrations with external APIs (Stripe, etc.).
• Orchestrating autonomous agents that require strict security and audit controls.
• Packing and supplying compact code context to agents for large monorepos.
• Performing high-speed codebase search and symbol mapping across millions of lines.
• Integrating Edge-first authentication flows using Auth.js v5 in modern deployments.

Best practices

• Use environment-driven model registries rather than hardcoding model names to simplify upgrades.
• Enforce strict-auditor checks before agents commit to main branches or deploy changes.
• Prefer async iterators for SDK pagination (for await...of) instead of manual starting_after handling.
• Keep JWTs lean; avoid storing PII and use AUTH_ prefixed secrets for v5 deployments.
• Create targeted repomix packs with explicit include/ignore lists to limit token usage.

Example use cases

• Auto-diagnose and remediate Stripe API version or schema errors with GPT-5 generated diagnostics and retry logic.
• Edge-first Auth.js v5 setup that validates AUTH_SECRET and enforces AUTH_TRUST_HOST for non-hosted environments.
• Generate a codebase map with repomix and ripgrep to produce a compact AI review bundle for a large refactor.
• Run automated agent audits that prevent unsafe changes to critical packages in a pnpm/Bun monorepo.
• Use high-performance search commands to find sensitive keys and enforce secret-remediation workflows before CI artifacts are published.