@getcolter/cli

@getcolter/cli

AI agent readiness platform for e-commerce. Check protocol support, fix gaps, monitor agent traffic, and test with AI personas.

Install

npm install -g @getcolter/cli

Or run directly:

npx @getcolter/cli check https://your-store.com

What it does

Command	Purpose
colter check	Instant readiness scan — protocols, structured data, agent surfaces, 5-dimension scoring
colter fix	Generate and apply fixes for agent readiness gaps
colter lens	Register a store for request-level AI agent traffic monitoring
colter test	Run hosted AI mystery shopper tests, or --local BYOK tests with your provider key
colter query-rank	Local/BYOK recommendation competitiveness test
colter ucp-qa	Local/BYOK UCP buyer-journey QA
colter batch	Scan multiple stores from a YAML file
colter mcp	Start MCP server for AI agent integration

Quick start

# Free readiness scan (no account required)
colter check https://your-store.com

# Detailed JSON output with scores
colter check https://your-store.com --json --verbose

# Generate a remediation plan
colter fix https://your-store.com

# Register a store for Lens monitoring and full-coverage setup
colter lens https://your-store.com

# Run hosted AI mystery shopper simulations
colter auth login
colter test https://your-store.com

# Run local/BYOK agent QA using your provider key
export GEMINI_API_KEY=...
colter query-rank https://your-store.com
colter ucp-qa https://your-store.com
colter test https://your-store.com --local

Protocols supported

Protocol	What it does	Agents using it
UCP	Full-lifecycle commerce (discover, browse, cart, checkout)	Gemini, AI Mode, Copilot, Shopping Graph
ACP	Agent checkout and payment	ChatGPT, Operator
MCP	Tool-based agent integration	Claude, MCP-compatible agents
WebMCP	Browser-native agent tools (W3C Draft)	Chrome flag, Chrome 149 Origin Trial, browser agents

Lens coverage

Lens is designed for request-level agent traffic capture. Deploy the Lens edge worker or configure server-log ingestion for full coverage, including direct crawlers that never execute JavaScript. The JavaScript beacon remains available as a fallback for browser-like agents only.

Scoring

Every check produces a composite score (0-100) across 5 dimensions:

• Discovery — Can agents find your store?
• Transaction — Can agents buy from your store?
• Security — Is the store secure for agent transactions?
• Ecosystem — How many AI platforms can access your store?
• Content Quality — Can agents understand your products?

MCP server

Colter includes an MCP server for AI agent integration:

{
  "mcpServers": {
    "colter": {
      "command": "colter",
      "args": ["mcp", "--admin-tools"]
    }
  }
}

Platforms

Pre-built binaries for macOS (ARM64, x64), Linux (x64, ARM64), and Windows (x64).

Links

• Documentation
• Pricing
• GitHub
• Report an issue

License

Colter is free to use. See LICENSE for terms.