adnanbdm555-wq

Brandloo AI Agent System — Testing skill for Claude Code

Testing community

Brandloop — Autonomous multi-agent social media operations platform built with Next.js 14, Drizzle ORM, PostgreSQL & Claude AI.

How to install Brandloo AI Agent System

This entry records only its repository, not the path inside it, so there is no exact command to give. Open adnanbdm555-wq/brandloo-ai-agent-system and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Brandloo AI Agent System does

Brandloop — Autonomous multi-agent social media operations platform built with Next.js 14, Drizzle ORM, PostgreSQL & Claude AI. Features chained Strategy, Content, Creative, Video, and QA pipelines with multi-tenant agency management.

Alternatives in Testing

Testing Strategy — This project is small, runs in a terminal, and is mostly deterministic 10.9k ★
OpenAgentsControl — AI agent framework for plan-first development workflows with approval-based execution 4.8k ★
Create Test Repo — Create a new ArcKit test repo with full scaffolding, auto-detecting the next version number 2.2k ★

README

Brandloop — Social Media Operations Platform

**All 6 phases complete, plus a full commercial billing system and a chained agent Pipeline on top.** Authentication, multi-tenant Agencies, Brand Management, Brand Knowledge Base, Campaigns, Strategy Agent, Content Agent, Content Calendar, Content Studio, Creative Agent, Video Agent, QA Agent, Approval System, Social Accounts, Publishing tracking, Notifications, Analytics, AI Insights & Optimization Agent, Templates, Automation Builder, Users & Roles, Billing (15-day trial, automatic invoicing, automatic suspension, Safepay checkout) — and now a **Pipeline**: Strategy → Content → Creative → Video → QA chained into one run, matching the original architecture diagram this was built from.

Built from the master prompt's architecture diagram (Strategy → Content → Creative → Video → QA → Approval → Publishing → Analytics → Optimization), phase by phase, with every feature actually run and tested against a real Postgres database before moving to the next — see "What's real vs. what needs your own keys" below for the handful of places that need credentials only you can provide.

What's working right now

**Pipeline**

One trigger runs Strategy → Content → Creative → Video → QA in sequence — the actual chain from the original architecture diagram this platform was built from, rather than five separate manual visits to five separate Studios.
Each step's outcome is recorded and shown — success, failed, or skipped (Strategy skips if the campaign already has one or none was selected; Video skips for non-video content types) — with an optional final "auto-submit for approval" step that only fires if QA passed.
A missing ANTHROPIC_API_KEY aborts the whole run immediately after the first step that needs it, with one clear message, instead of failing the same way five times in a row. Tested directly: with no campaign selected the run correctly skips Strategy and fails cleanly at Content; with a campa