No CMS or framework detected, hosted on Fly.io. Domain registered in 2026.
nothumansearch.ai is hosted on Fly.io. It sends email through Amazon SES.
| Technology | Version | First seen |
|---|---|---|
| CDN & Hosting1 | ||
| – | 30 Oct 2026 | |
| Transactional Email2 | ||
| – | 30 Oct 2026 | |
| – | 30 Oct 2026 | |
| Fonts & Icon Fonts1 | ||
| – | 30 Oct 2026 | |
Registration from the registry's RDAP record, the certificate the site serves, and its rank in public traffic lists.
The crew hasn’t read this site’s colors yet. We read a palette when we capture a site’s screenshot, busiest sites first. Meanwhile, browse palettes from other sites.
inbound-smtp.us-east-1.amazonaws.comns35.domaincontrol.comns36.domaincontrol.comv=DMARC1; p=quarantine; adkim=r; aspf=r; rua=mailto:dmarc_rua@onsecureserver.net;# Not Human Search — robots.txt # We index sites for AI agent discovery. We welcome all crawlers. User-agent: * Allow: / # AI agent crawlers — explicitly welcome (matches 8bc coverage, 2026-04-15) User-agent: GPTBot Allow: / User-agent: ChatGPT-User Allow: / User-agent: OAI-SearchBot Allow: / User-agent: ClaudeBot Allow: / User-agent: Claude-Web Allow: / User-agent: anthropic-ai Allow: / User-agent: PerplexityBot Allow: / User-agent: Google-Extended Allow: / User-agent: Applebot-Extended Allow: / User-agent: Meta-ExternalAgent Allow: / User-agent: FacebookBot Allow: / User-agent: CCBot Allow: / User-agent: Bytespider Allow: / User-agent: Amazonbot Allow: / User-agent: cohere-ai Allow: / User-agent: Diffbot Allow: / User-agent: YouBot Allow: / User-agent: DuckAssistBot Allow: / User-agent: PetalBot Allow: / User-agent: FirecrawlAgent Allow: / Sitemap: https://nothumansearch.ai/sitemap.xml # RSS feed: https://nothumansearch.ai/feed.xml (new agent-ready additions) # AI agent discovery # llms.txt: https://nothumansearch.ai/llms.txt # llms-full.txt: https://nothumansearch.ai/llms-full.txt # OpenAPI spec: https://nothumansearch.ai/openapi.yaml # AI plugin manifest: https://nothumansearch.ai/.well-known/ai-plugin.json # MCP manifest: https://nothumansearch.ai/.well-known/mcp.json # MCP endpoint (streamable-http JSON-RPC): https://nothumansearch.ai/mcp # Registry auth proof: https://nothumansearch.ai/.well-known/mcp-registry-auth # Security contact: https://nothumansearch.ai/.well-known/security.txt
# Not Human Search
> The Google for AI agents. Find any service, API, or tool — ranked by agentic readiness.
Not Human Search is an agentic-first search engine. When your AI agent needs to discover a payment API, a job board, an ecommerce platform, or any web service, NHS returns results ranked by how well each site serves non-human users.
Site owners: score your own site's agent-readiness (0-100, 7 signals) and monitor it over time, so AI agents can actually find and use you.
We index 6154+ sites and score them 0-100 based on 7 agentic signals: llms.txt, ai-plugin.json, OpenAPI specs, structured APIs, MCP servers, robots.txt AI rules, and Schema.org markup.
## MCP Server (preferred for agents)
Not Human Search is itself an MCP server. Wire it into your agent once and get live agentic-web search as a first-class tool.
Endpoint: https://nothumansearch.ai/mcp
Transport: streamable-http
Tools (14): search_agents, get_site_details, get_stats, list_categories, get_top_sites, submit_site, register_monitor, verify_mcp, find_mcp_servers, recent_additions, check_url, record_action_interest, prepare_provider_action, handoff_provider_action
Claude Code setup:
claude mcp add --transport http nothumansearch https://nothumansearch.ai/mcp
## Quick Start — Search the Agentic Web
GET https://nothumansearch.ai/api/v1/search?q=payment+API
GET https://nothumansearch.ai/api/v1/search?q=AI+jobs
GET https://nothumansearch.ai/api/v1/search?q=ecommerce+api
GET https://nothumansearch.ai/api/v1/search?q=weather+data
GET https://nothumansearch.ai/api/v1/search?q=authentication
## API Reference
Base URL: https://nothumansearch.ai/api/v1
### Search
GET /search?q={query}&category={cat}&tag={tag}&min_score={0-100}&has_api=true&has_mcp=true&has_openapi=true&has_llms_txt=true&page={n}
Returns: {results: [...unchanged neutral organic results...], paid_offers: [...separate disclosed provider-funded actions...], total, page, has_next, search_id}
### Optional Provider-Funded Actions
Search and canonical provider links remain free. An active provider offer can
appear only in the separate paid_offers collection beside a site already returned
organically. Money never changes organic membership, score, order, or total.
Each public offer sidecar exposes offer_version and the exact
commercial_terms_contract_version/commercial_terms_sha256, fixed credit rule,
provider response expectation, first-activation terms anchor rule, and provider
Merchant-of-Record acknowledgement used for that version.
POST /api/v1/action-tickets
Creates a ticket after the caller supplies search_id, offer_id, one controlled
demand topic already on the receipt, and the exact versioned principal-
authorization attestation. It returns the raw attribution_token once (or
reconstructs it for an exact replay) plus the POST handoff_endpoint; it does not
return the provider action URL. Creating a ticket charges neither party.
POST /api/v1/action-tickets/handoff
Body: {"ticket_id":"...","attribution_token":"...",
"principal_handoff_consent":true,
"handoff_consent_version":"nhs-provider-handoff-consent-v1"}
Presents the ticket bearer and separate handoff-time principal attestation only
in a JSON body; every response is private, no-store. NHS first commits one
append-only nhs-action-handoff-v1 receipt bound to the exact ticket, offer
version, commercial-terms hash, and handoff-consent version, then returns the
attributed provider action URL. The receipt contains a one-way hash of the
presented bearer but no query, agent/principal identity, contact data, network
address, referrer, or user agent. The handoff charges neither party; only the
disclosed authenticated provider-reported downstream event can create a
provider charge.
POST /api/v1/provider/commercial-acceptances
Headers: X-NHS-Provider-Key and Idempotency-Key
Records a provider-authenticated pilot-company, exact-terms acceptance, or
exact-terms renewal. A terms event must carry the exact offer_version and
exact_terms_sha256 the provider reviewed; a changed draft is rejected rather
than silently accepted. This provider event is append-only but is not funding,
owner verification, or commercial proof by itself; the keyed company and exact
commercial evidence are verified separately by the NHS owner.
Consent wording: /privacy#consent-v1
Handoff consent wording: /privacy#handoff-consent-v1
Provider setup: /providers
### Caller-Attested Action Interest (No Provider Contact)
POST /api/v1/action-interests
After an organic search, a caller can record that its human or company principal
currently wants one controlled next step with an exact returned domain: quote,
trial, demo, booking, application, signup, or purchase. This private receipt
expires with the source search, no later than 30 days after that search, and is
used only for aggregate Stage 1 demand. It does not contact the
provider, create a provider action ticket or charge, affect rank or score, or
count as commercial proof. Exact wording: /privacy#action-interest-v1
### Site Details
GET /site/{domain}
Returns: full site profile with llms.txt content, OpenAPI summary, all signals
### Submit a Site
POST /submit Body: {"url": "https://example.com"}
We crawl immediately and add it to the index.
### On-Demand Check (CI / pre-deploy)
POST /api/v1/check Body: {"url": "https://example.com"}
Returns live agentic readiness score without waiting for the crawl queue.
Free tier: 10 checks/hour per IP. Great for CI pipelines that fail the build
when a site's agent signals regress.
### Optional Priority API Key
Search, site details, and organic results are free without a key. An active key
only raises hourly safety ceilings; it never changes results, rank, or score.
GET /api/v1/api-keys/subscribe
Returns the optional priority-throughput plan and the machine contract for creating a Stripe Checkout session.
POST /api/v1/api-keys/subscribe Body: {"email": "you@example.com", "plan": "unlimited"}
Plan: $9.99/month for 50,000 priority-throughput calls. After that allocation,
requests continue at free safety limits instead of returning a payment error.
Returns: {checkout_url, plan, monthly_limit, amount_cents, activation_url}
### Stats
GET /stats
### Top Sites
GET /top?has_mcp=true&category=ai-tools&limit=25
Returns the highest-scored sites in the index, sorted by agentic_score DESC.
Designed for embedding / mirroring: stable URL, cached 5min, no auth. Max 100 results.
### Categories
GET /categories
Returns: {categories: [{name, count}]} — live public categories plus audit-only buckets when present.
### Monitor a Site
POST /monitor/register Body: {"email": "you@x.com", "domain": "site.com"}
Email alert when a site's agentic readiness drops. Returns an unsubscribe URL.
Free tier: multiple monitors per email allowed, one per domain.
## Categories
Public categories: ai-tools, developer, data, finance, ecommerce, jobs, security, health, education, communication, productivity, news.
Audit-only buckets: other, spam. These are exposed for transparency and filtering, not promoted as agent-ready inventory.
## Scoring (0-100)
- llms.txt: 25 pts
- ai-plugin.json: 20 pts
- OpenAPI spec: 20 pts
- Structured API: 15 pts
- MCP server: 10 pts
- robots.txt AI rules: 5 pts
- Schema.org: 5 pts
## Make Your Site Agent-Ready
Step-by-step recipes for each of the 7 signals — copy-paste examples for llms.txt, ai-plugin.json, OpenAPI, MCP, and more.
Guide: https://nothumansearch.ai/guide
Live scorer: https://nothumansearch.ai/score
Report: https://nothumansearch.ai/report
## Links
- Search: https://nothumansearch.ai/api/v1/search?q=
- MCP Server Directory: https://nothumansearch.ai/mcp-servers
- Full Index: https://nothumansearch.ai/llms-full.txt
- OpenAPI: https://nothumansearch.ai/openapi.yaml
- Plugin: https://nothumansearch.ai/.well-known/ai-plugin.json
## Developer Resources
- One-line install (Claude Code / Cursor / Cline / Continue):
curl -fsSL https://nothumansearch.ai/install | sh
- Verify any MCP server in 3 curls: https://gist.github.com/unitedideas/ce709323717b95eb56f7be7392a0a557
- Q2 2026 agent-ready sites curation (120 categorized): https://gist.github.com/unitedideas/c60bb35943ef609f99123bdfae146e55
- NHS Score Check GitHub Action (fail CI on score drop): https://github.com/unitedideas/nhs-score-check-action
- Q2 2026 MCP Ecosystem Health data + methodology: https://8bitconcepts.com/research/q2-2026-mcp-ecosystem-health.html
Contact: mailto:security@nothumansearch.ai Contact: https://github.com/unitedideas/nothumansearch/issues Expires: 2027-11-12T09:32:28Z Preferred-Languages: en Canonical: https://nothumansearch.ai/.well-known/security.txt Policy: https://github.com/unitedideas/nothumansearch/blob/main/SECURITY.md # Not Human Search — security contact # Report vulnerabilities privately via the Contact addresses above. # Please do not test authenticated flows against production — our admin # endpoints rate-limit + log per-IP; accidental DoS triggers paging. # Responsible-disclosure window: 30 days before public write-up.