---
spec: two-tier/2.0.0
schemaVersion: "https://mahoneycontextprotocol.com/specs/two-tier/v2"
date_modified: "2026-06-11"
facts:
  name: Is It Agent Ready?
  email: "joe@joemahoney.io"
  portfolio: "https://mahoneycontextprotocol.com"
  description: Scores any website for AI-agent readiness on a 100-point navigability scale with a public leaderboard and remediation playbook
  intake: Static site with client-side JSON tools (scan_site, get_leaderboard); optional live backend when SCANNER_ENDPOINT is configured
---

# Is It Agent Ready?

Scores any website for AI-agent readiness on a 100-point navigability scale.
Live scanner, public leaderboard, and a remediation playbook. By
MahoneyContextProtocol.

## The web is being rebuilt for agents. Most sites are invisible to them. {#hero}

Point our scanner at any URL. Get a 100-point agent-navigability score, a
per-check breakdown, and exactly what to ship to fix it. Then see where you
land on the public leaderboard.

**Current stats** (from `/data/leaderboard.json`):
- Sites scanned: 54
- Average score: 33/100
- Agent-ready (≥70): 1
- Forfeited agent-channel opportunity: $22.4M

## Three layers, one verdict {#layers}

Agent-ready isn't one thing. A site has to be legible to machines, operable by
agents, and still excellent for humans. The scanner checks all three.

| Layer | Focus | What it checks |
|---|---|---|
| Machine-legible | The rubric | robots content-signals, llms.txt, JSON-LD, API catalog, OAuth discovery, MCP cards — the standards agents actually read |
| Agent-operable | The pull model | Agent Skills, A2A cards, WebMCP tools, signed-bot auth — surfaces that let an agent *do* something, not just read |
| Human craft | Still a site | The score rewards navigability and coherence: cross-references that resolve, handshakes that complete, digests that match |

## The leaderboard is the product {#leaderboard-preview}

Seeded with an empirical 53-site benchmark of major platforms — none of them
ready. One engine-built site holds the only live agent-ready row. Live scans
write new rows.

**Top 6** (from `/data/leaderboard.json`):
1. landscapingnear.me — 81 (Agent-Ready, live scan)
2. doit.com — 58 (Partial, benchmark)
3. monday.com — 58 (Partial, benchmark)
4. asana.com — 53 (Partial, benchmark)
5. cohere.com — 53 (Partial, benchmark)
6. freshworks.com — 53 (Partial, benchmark)

Full leaderboard: [/leaderboard](/leaderboard)

## How it works

The scanner runs a 100-point agent-navigability probe:
- **Surface coverage (40 pts)**: robots.txt, llms.txt, JSON-LD, MCP card, A2A card, API catalog, agent skills, WebMCP, OAuth discovery, security.txt, humans.txt
- **Edge resolution (30 pts)**: cross-references between surfaces resolve, digests verify, anchors land on stable ids
- **Handshake depth (20 pts)**: OAuth flows complete, MCP tools validate as JSON Schema, A2A skills have input+output schemas
- **Coherence (10 pts)**: facts byte-identical across JSON-LD + agent YAML + markdown twin frontmatter

Every scan issues a sha256 provenance receipt that re-verifies at output time.

## Agent tools {#tools}

Invokable via WebMCP (`navigator.modelContext`, registered by `/webmcp-init.js`)
and listed as A2A skills in `/.well-known/agent.json`:

- `scan_site({ url })` — Score a URL for AI-agent readiness (0-100) with a
  per-check breakdown, remediation guidance, and a sha256 provenance receipt.
  Static-fallback: reads `/data/proof.json` for baked scan results when no live
  backend is configured; optionally calls `SCANNER_ENDPOINT` if available.
  Returns `{domain, score, verdict, checks, remediation, receipt}`.

- `get_leaderboard()` — Return the public agent-readiness leaderboard with
  stats, ranked rows, and the field's most common failures. Static-fallback:
  reads `/data/leaderboard.json`. Returns `{stats, rows, common_failures,
  generated_at}`.

## Agent surfaces {#surfaces}

- llms.txt: /llms.txt
- A2A card: /.well-known/agent-card.json
- Agent card: /.well-known/agent.json
- MCP card: /.well-known/mcp/server-card.json
- API catalog: /.well-known/api-catalog
- Agent skills: /.well-known/agent-skills/index.json
- OpenAPI: /openapi.json
- WebMCP: /webmcp-init.js
- JSON-LD: in-page application/ld+json
- Operating manual: /operate.txt (if present)
- HTML page: https://itisagentready.com/ (human layer with HUMAN/AGENT toggle)
- Markdown twin: /index.html.md (this file)

Conformance: two-tier/2.0.0, conneg: static-fallback (static host; this twin
plus llms.txt is the declared reader-surface delivery — the host additionally
attempts header negotiation of this twin where Apache rewrite is honored).

## Contact

- Email: joe@joemahoney.io (`mailto:joe@joemahoney.io`)
- Part of the MahoneyContextProtocol portfolio: https://mahoneycontextprotocol.com
- Scanner engine: 100-point agent-navigability probe (open-source reference implementation)

## The scoring model

Surface coverage (40) + edge resolution (30) + handshake depth (20) + coherence
(10). A high score means an agent can land on the homepage, discover every
surface via Link headers, follow each cross-reference until it resolves, and
have digests/anchors cross-validate. The engine is the agent-readiness
navigability probe.

## Note

itisagentready.com (IT-IS) is the MahoneyContextProtocol product. Not
affiliated with any similarly named third-party scanner.