index-ai-validator

Appearance

index-ai-validatorIs your site readable by AI agents?

A free CLI that checks whether your site exposes a clean, agent-facing layer — index-ai manifest, Agent Index, clean endpoints, and measured content size. Runs in your terminal. No signup.

Get started
Fix your report

Prove agents can find you

Checks the AI Manifest at /.well-known/index-ai.json and the discovery hints agents rely on — HTML link, HTTP Link header, robots.txt, llms.txt.

Validate your Agent Index

Inspects the Level 2a graph nodes, fetches each clean llm_url endpoint, and verifies content_chars against the Unicode NFC code-point count.

Catch leaks before agents do

Flags secret-shaped values, sensitive variable names, and private infrastructure references in the public content you expose to agents.

Most sites are readable by browsers. Is yours readable by agents? ​

Browsers read HTML, CSS, and JavaScript. AI agents need a different interface: a clean layer that says what a site is, where its content lives, how fresh it is, and how much text they will pay tokens for before fetching it.

index-ai explores an Agent View through three ideas — an AI Manifest that describes the site, an Agent Index that maps public content into structured nodes, and clean content endpoints that return Markdown or plain text instead of rendered HTML.

@hardmachinelabs/index-ai-validator makes that layer testable. One command tells you whether yours works.

Run it ​

bash
npx @hardmachinelabs/index-ai-validator https://example.com

The package name is @hardmachinelabs/index-ai-validator. The CLI binary is index-ai. By default it prints a deterministic, summary-first report:

txt
index-ai validation result

Target: https://example.com
Duration: 42 ms
Conformance: level-2a
Passed: true

Summary:
- pass: 12
- warn: 0
- fail: 0
- total: 12

Metrics:
- manifest_found: true
- agent_index_found: true
- total_nodes: 6
- valid_clean_endpoints: 6
- valid_content_chars: 6

No failures or warnings.

Next:
- No blocking validation fixes were found.

Add --json for a stable machine-readable result, or --html report.html for a shareable visual report with a CI verdict, a readiness score, and recommended next steps.

Checks Level 1 + Level 2a today. Not certification, not a traffic promise. → See the full scope

What it checks ​

  • Level 1 AI Manifest: fetch, JSON content type, JSON parse, and schema shape
  • The access.agent_index declaration and the Agent Index graph it points to
  • Level 2a node fields, llm_url structure, and clean endpoint content types
  • Hard HTML leaks, with tolerated soft inline markup reported as warnings
  • content_chars in exact and max modes, using Unicode NFC code-point counting
  • Secret-shaped values and private infrastructure references in public AI-facing content
  • Discovery hints on the homepage, robots.txt, and /llms.txt

For what it deliberately does not do, see Scope.

Free tool. Need the layer built for you?Get your agent-facing layer reviewed and shipped by the maker of index-ai.
AI-readable auditContact