Year · 2026–presentBy Ayush Niroula
Self-hosted AI market research and marketing simulation built for AI agents. MarketMirror interviews LLM buyer personas, places your ads among ordinary posts in simulated Facebook, Instagram, Reddit, TikTok, LinkedIn and X feeds, lets buyers argue in a CAMEL-AI OASIS social simulation, runs surveys, email and sales-pitch tests, and sends simulated visitors through a real website. Ask in plain words from Claude Code, Codex, Gemini CLI, Cursor or any MCP client and get a go / iterate / no-go report in one offline HTML dashboard. Runs can mix model providers to expose where they disagree, and real campaign exports can be paired with simulated outcomes to check how well the simulation held up. Built on MiroFish and CAMEL-AI's OASIS; AGPL-3.0.
Highlights
- Live demo dashboard from a 2,999-buyer simulated run, with buyers, objections, pricing, funnel and per-model results
- Listed in the official MCP Registry (io.github.DarkWorldCoder/market-mirror) and installable with uvx marketmirror
- Buyer personas carry budgets, devices, trust concerns and decision cards, plus memory across tests in the same run
- Guided local app with spend limits reviewed before each paid stage; only model calls leave the machine
Tech stack
Python
Flask- CAMEL-AI OASIS
MCP
OpenAI- Anthropic
Vue
D3
Playwright
Docker