Two-minute walkthrough
From install to first citation
in two minutes.
Watch Canonry go from a fresh terminal to a multi-provider AEO citation sweep across Gemini, OpenAI ChatGPT, Anthropic Claude, and Perplexity. No edits, no cuts — this is the actual run, on a local machine, against local SQLite.
What you'll see
Four moves. Forty lines of CLI.
Project setup
Create the canonry project, pointing at the domain you want to track. The CLI scaffolds local SQLite storage at ~/.canonry/data.db.
cnry project create ainyc --domain ainyc.ai --country US --language enQueries and competitors
Register the phrases you care about being cited for, then the competitors you want to benchmark against.
cnry query add ainyc "best AEO agency NYC" "AI SEO agency NYC"First multi-provider sweep
Trigger a sweep against Gemini — the same syntax works for openai, claude, and perplexity. Each provider is asked your phrases and the answer text is captured locally.
cnry run ainyc --provider geminiInspect citations
Open the dashboard at localhost:4100 to see mention rates, per-phrase citation history, the answer text each provider returned, and which competitor sites got mentioned alongside you.
open http://localhost:4100
Bundled agent
Aero is the cat in the bottom-right.
The Aero agent ships in the box. After every sweep it surfaces the diff — what you gained, what you lost, which competitors moved — and offers concrete next moves (page edits, schema changes, internal links). You can also bring your own coding agent: Claude Code, OpenAI Codex, Cursor, OpenClaw, or GitHub Copilot, connected over the canonry CLI or MCP.
Run it yourself
Same demo. One prompt.
Paste the canonry setup prompt into Claude Code or OpenAI Codex and your agent installs canonry, scaffolds config, installs the bundled canonry and aero skills, and runs your first citation sweep — same flow as the video, end to end.
npm install -g @ainyc/canonry