Skip to main content

For Developers

Five tools. One CLI.

Check readiness, generate fixes, test with AI personas, monitor agent traffic, and verify conformance. CLI, API, and MCP.

Install the CLIAPI reference
terminal
$ npm install -g @getcolter/cli
$ colter check my-store.com
Scanning protocols, structured data, agent surfaces...
✓ UCP manifest valid
✓ ACP checkout endpoint found
✗ Missing llms.txt
✗ No WebMCP tools detected
 
Score: 64/100 (C) — 4 fixes available
$ colter fix my-store.com --apply
✓ Applied 4 fixes — re-scan score: 91/100 (A)

The product suite

Each product is a CLI subcommand, an API endpoint, and an MCP tool.

Check

Free

Free readiness scan. 5-dimension scoring across 30+ signals. Instant results via CLI, API, or web.

$ colter check my-store.com

Fix

Generate platform-specific protocol fixes. View plans on free, auto-apply on Pro.

$ colter fix my-store.com --apply

Lens

Real-time AI agent traffic monitoring. See which agents visit, where they drop off, and revenue impact.

$ colter lens status --site my-store.com

Test

AI persona testing. 10 personas across Claude, GPT, and Gemini attempt real shopping flows.

$ colter test my-store.com --browser

Verify

Evidence Packs for conformance. Cryptographically signed proof that protocols are implemented correctly.

$ colter verify my-store.com

Three ways to integrate

Same data, same scores, same fixes. Pick the interface that fits your workflow.

CLI

Install globally, run from anywhere. All five products available as subcommands.

$ npm install -g @getcolter/cli

REST API

Programmatic access. POST a URL, get structured results. API keys on Pro and Agency.

$ POST https://agenticcom.ai/api/v1/check

MCP Server

Plug Colter into Claude, Cursor, or Windsurf. Your AI agent can check and fix stores directly.

$ colter mcp

Five protocols. Full coverage.

Colter checks, fixes, tests, and monitors across every major agentic commerce protocol.

UCP
Google

Universal Checkout Protocol for product discovery and transactions

ACP
OpenAI

Agent Commerce Protocol for ChatGPT and GPT-powered agents

MCP
Anthropic

Model Context Protocol for Claude and tool-use agents

A2A
Google

Agent-to-Agent protocol for multi-agent collaboration and task delegation

WebMCP
W3C

Browser-native tool registration via HTML form attributes

CI/CD integration

Gate deployments on readiness score. Catch protocol regressions before they reach production.

.github/workflows/readiness.ymlyaml
name: Agent Readiness Check
on: [push, pull_request]
 
jobs:
  readiness:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Install Colter
        run: npm install -g @getcolter/cli
      - name: Check readiness
        run: colter check ${{ secrets.STORE_URL }} --ci --threshold 60
      - name: Run agent tests
        run: colter test ${{ secrets.STORE_URL }} --json
      - name: Upload evidence
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: evidence-pack
          path: ./evidence/

Pricing that scales with you.

Free

$0

Get started. No account required.

  • 3 checks per day
  • Fix plans (view only)
  • 24-hour Lens window
  • CLI + web access
Try it now

Pro

$49/mo

Full toolkit for your projects.

  • 50 checks per day
  • Fix auto-apply
  • Test: 10 personas, 3 LLMs
  • 30-day Lens history + AI insights
  • Verify: Evidence Packs
  • CLI + API + MCP access
Start free trial

7-day free trial, no credit card required.

Agency

$199/mo

Portfolio tools for client work.

  • Everything in Pro
  • 500 checks/day, 100 stores
  • Portfolio dashboard
  • White-label reports
  • Batch operations
  • JSON API export
Start with Agency

Common questions

How do I install and run my first check?

Install the CLI with npm install -g @getcolter/cli, then run colter check your-store.com. No account needed. You'll get a readiness score, protocol breakdown, and fix recommendations in about 10 seconds.

How does MCP integration work?

Run colter mcp to start a local MCP server. Point your AI agent (Claude, Cursor, Windsurf) at it. The agent can then call check, fix, and verify as tools — useful for AI-assisted development where the agent checks readiness as part of its workflow.

Can I add Colter to my CI/CD pipeline?

Yes. Add colter check $STORE_URL --ci --threshold 60 to your GitHub Actions, GitLab CI, or any pipeline. The --ci flag exits non-zero when the score drops below the threshold, so you can gate deploys on readiness regressions. Add --json for machine-readable output.

What data does Check access?

Check is read-only and examines only publicly available data: HTML, JSON-LD, robots.txt, sitemaps, protocol manifests, and HTTP headers. It never authenticates, never writes, and never accesses admin panels or customer data.

How do I install Lens on my store?

Add a small JavaScript snippet to your site — similar to Google Analytics. On Shopify, you can use the Lens theme block instead. Once installed, agent traffic appears in your Lens dashboard within minutes. Free tier gives a 24-hour data window; Pro gives 30 days.

Start with a scan. Build from there.

Install the CLI, check a store, see the score. Everything else follows.

npm install -g @getcolter/cli
colter check your-store.com
Check a store nowRead the docs