I automated competitor research with a team of AI agents (FREE REPO INCLUDED)

GET THE FREE REPO (install it for your own company): https://github.com/sabahudin-web/comp... Bright Data powers the live web data the scouts pull from. Get $20 of free credit (enough to run this for a month) with my link: https://brdta.com/sabahudin (promo code: sabahudin) Subscribe to my newsletter: https://sabahudinmurtic.substack.com/ --- In this build I wire up a competitor research system in Claude Code: six AI agents that watch your competitors every week, with a source link on every claim. Most AI just answers questions. This one does a real job end to end while you watch. We build a Claude Code plugin that runs two engines: a team of parallel scout agents (one AI agent per competitor) that pull live web data with Bright Data MCP, and a strategic "war room" of analyst agents that debate the findings and agree on what to do next. Every signal carries a verifiable source URL. It lands in three places you own: local files, a shareable Notion database, and a clean HTML dashboard you open in your browser. Built live from an empty folder to a working tool in about 20 minutes. If you run a consulting or solo business and want competitive intelligence that updates itself, this is the workflow. The full plugin is a free repo, link below, so you can install it for your own company and point it at your competitors. P.S. I also show how to package the agent team as a plugin and push it to a GitHub repo so you can resell it to clients. Timestamps (chapters, start at 0:00): ``` 0:00 What 6 agents watching your competitors looks like 0:22 The two engines: parallel scouts + a debate war room 2:18 What we are building (and why every claim is sourced) 2:39 Setup: verify Bright Data and Notion MCP first 4:07 Engine 1: parallel scout agents, one per competitor 4:41 The Bright Data tools each scout pulls 5:31 Sub-agents vs real agent teams (when to use each) 6:06 Engine 2: the strategic war room of analysts 6:57 Three outputs: files, Notion, HTML dashboard 8:13 Building it live: the starting prompt 10:28 Create the folder and plan the build 14:15 Reviewing the plan 15:49 Phase 1 build: rules, references, agents 17:14 First run: entering your business and competitors 19:05 The live dashboard and cited signals 21:09 The war room debate and recommendation 21:50 The same output inside Notion 23:34 Package it as a plugin and ship the GitHub repo 25:08 Wrap-up and the free Bright Data promo ``` #claudecode #aiagents #competitiveintelligence