The Trust Layer API.

⚡ Quickstart — one call, full verdict

curl https://trustifer.com/api/v1/trust/0xd8dA6BF26964aF9D7eEd9e03E53415D37aA96045

{
  "verdict": "PROCEED",                 // machine-actionable
  "status": "active",
  "trust":  { "score": 8.9, "scale": "1-10", "band": "Strong", "confidence": "High" },
  "bureau": { "consensus": 80, "band": "Exceptional",
              "corroboration": "corroborated", "independentGroups": 4,
              "reasonCodes": ["Corroborated by 4 independent source groups …"] },
  "euAiAct": { "risk": "minimal", "label": "Minimal risk" },
  "links":  { "badge": "…/api/badge/0xd8dA…", "passport": "…/passport/0xd8dA…" }
}

Works with any EVM address or agent domain. Sanctioned / killed / blacklisted subjects return DO_NOT_ENGAGE — safety findings are never averaged away.

🛡 Trust Firewall verify-before-execute, enforced — our flagship

Don't just look up trust — enforce it on every action. One POST before each payment/hire/data-share returns a deterministic, credit-aware decision with an Ed25519-signed receipt for your audit trail. Sanctioned, blacklisted, or killed counterparties are BLOCKED at the floor — no policy can override safety.

curl -X POST https://trustifer.com/api/v1/firewall/check \
  -H "Content-Type: application/json" \
  -d '{"subject":"0xCounterparty…","action":"payments.send","amountUsd":400}'

{
  "decision": "LIMIT",            // ALLOW · LIMIT · REVIEW · BLOCK
  "limitUsd": 250,                // credit-aware ceiling (card-network style)
  "reasons": ["Consensus 58 in limit band (<60) — authorized up to $250",
              "Thin credit file — tier tightened by 10"],
  "consensus": 58, "status": "active",
  "signature": "ed25519…",        // verify: POST /api/v1/firewall/receipt
  "expiresAt": "…+5min"           // decisions are point-in-time
}

Tighten (never loosen) the default policy per call: policy: { blockBelow, limitUsd, requireApprovalAboveUsd, treatUnknownAs: "block" }. Free in preview · metered keys + org policies at launch.

📚 Endpoints keyless preview works · get a free key for higher limits

🤖 Trustifer for Agents — pay-per-call x402 · premium pulls, no account

Autonomous agents pay a few cents in USDC on Base (gasless, via the open x402 standard) to pull a premium product before they transact — no signup, no key. Basic trust/firewall checks stay free; only premium pulls charge. Trustifer is its own x402 facilitator (neutral, no third party).

# 1. Call a premium endpoint → HTTP 402 + payment requirements
curl https://trustifer.com/api/v1/diligence/0xAGENT
# → 402 { x402Version, accepts:[{ scheme:"exact", network:"base",
#         asset:USDC, maxAmountRequired:"500000", payTo, … }] }

# 2. Pay with any x402 wallet (gasless EIP-3009), retry with the header:
curl https://trustifer.com/api/v1/diligence/0xAGENT \
  -H "X-PAYMENT: <signed-payment>"
# → 200 signed dossier + X-PAYMENT-RESPONSE (settlement proof)

Discover priced endpoints + schema at /.well-known/x402. Pricing: deep due-diligence $0.50 · signed credit report $0.25 · basic trust/firewall free. Works with any x402 wallet (Bankr, Coinbase…). Prefer a plan? Use a metered API key instead. (Charging activates at launch — free in preview today.)

🔔 Signed webhooks — know the moment trust changes

curl -X POST https://trustifer.com/api/v1/watch \
  -H "Content-Type: application/json" \
  -d '{"subject":"0xYourAgent…","url":"https://your-app.com/hooks/trustifer"}'

# → { "id":"wch_…", "secret":"whsec_…" }   ← store the secret (shown once)

Every delivery is signed: X-Trustifer-Signature: sha256=HMAC-SHA256(secret, rawBody). Fires on ≥5-point consensus moves or any status change (flagged · blacklisted · killed). Checks run daily in preview — higher frequencies ship with API tiers at launch.

🛡 Speaks your standards OWASP LLM Top-10 · NIST AI RMF · MITRE ATLAS

Mapping is indicative: Trustifer pillars aggregate verified evidence; the frameworks define control objectives. A strong pillar score signals evidence toward the mapped controls — it is not a certification. Full scoring math, hard rules & dispute process: /methodology · challenge any score: /disputes.

🔌 One passport, any rail

🤖 MCP server — trust inside your agent

Give Claude Code, Cursor, or any MCP host three tools — get_trust_verdict · get_bureau_assessment · watch_agent — so your agent verifies counterparties before executing. Zero dependencies.

# Claude Code
claude mcp add trustifer -- node sdk/mcp/server.mjs

# any MCP config
{ "mcpServers": { "trustifer": { "command": "node", "args": ["sdk/mcp/server.mjs"] } } }

LangChain / CrewAI: call GET /api/v1/trust/{subject} from a custom tool — one HTTPS request, no auth needed in preview. npm package & hosted MCP endpoint: at launch.