nerq api

5M+ AI assets indexed. Free, no auth. All responses JSON.

quick start · kya · weekly · verified · benchmark · search · stats · discover · badges · mcp

SDK

pip install nerq

from nerq import NerqClient
client = NerqClient()
r = client.preflight("langchain")
print(r.trust_score, r.recommendation)  # 82.4 PROCEED

PyPI · Swagger UI

Quick Start

# ecosystem stats
curl https://nerq.ai/v1/agent/stats

# search for coding agents
curl "https://nerq.ai/v1/agent/search?q=code+review&min_trust=50"

# benchmark a category
curl https://nerq.ai/v1/agent/benchmark/coding

# KYA — Know Your Agent
curl https://nerq.ai/v1/agent/kya/langchain

KYA — Know Your Agent

Public due diligence. Trust score, compliance, risk level, verdict.

GET /v1/agent/kya/{name_or_uuid}

curl https://nerq.ai/v1/agent/kya/langchain

{
  "agent_name": "langchain",
  "trust_score": 72.3,
  "compliance_score": 65.0,
  "risk_level": "TRUSTED",
  "eu_risk_class": "limited",
  "days_active": 180,
  "stars": 12500,
  "verdict": "Indexed 180 days, trust 72/100 [TRUSTED]."
}

GET /kya

Interactive search page: nerq.ai/kya

Weekly Signal

Ecosystem snapshot: top agents, trust changes, framework adoption, categories.

GET /v1/agent/weekly

curl https://nerq.ai/v1/agent/weekly

{
  "week_of": "2026-03-09",
  "new_indexed_count": 2399,
  "active_this_week": 4252,
  "ecosystem": {"total_agents": 66096, "total_tools": 60101, "total_mcp": 17468, "avg_trust_score": 67.3},
  "top_agents": [...],
  "agent_of_the_week": {"name": "...", "trust_score": 92.9, "grade": "A+"},
  "trust_changes": [...],
  "trending_frameworks": [...],
  "top_categories": [...]
}

GET /weekly

HTML page: nerq.ai/weekly

Nerq Verified

Agents with trust score ≥ 70. Verified status shown on KYA and agent pages.

GET /v1/agent/verified

curl https://nerq.ai/v1/agent/verified

{
  "verified_count": 18460,
  "threshold": 70,
  "agents": [
    {"name": "...", "trust_score": 92.9, "grade": "A+", "badge_url": "https://nerq.ai/badge/..."},
    ...
  ]
}

GET /verified

Browse all verified agents: nerq.ai/verified

Benchmark

Ranked leaderboards by category. Top 20 by trust score.

GET /v1/agent/benchmark/categories

curl https://nerq.ai/v1/agent/benchmark/categories

[
  {"category": "coding", "count": 10939, "avg_trust_score": 67.7},
  {"category": "security", "count": 1160, "avg_trust_score": 68.5},
  ...
]

GET /v1/agent/benchmark/{category}

Response header X-Total-In-Category has the full count.

curl https://nerq.ai/v1/agent/benchmark/coding

[
  {
    "agent_name": "ccmanager",
    "trust_score": 90.9,
    "compliance_score": 87.0,
    "risk_level": "TRUSTED",
    "days_indexed": 28,
    "platform": "github",
    "github_stars": 831
  },
  ...
]

Search

Fulltext search across 204K agents, tools, MCP servers.

GET /v1/agent/search

curl "https://nerq.ai/v1/agent/search?q=code+review&type=agent&min_trust=50&limit=5"

{
  "results": [
    {"name": "code-review-agent", "agent_type": "agent",
     "trust_score": 71.5, "category": "coding"},
    ...
  ],
  "total": 142,
  "limit": 5,
  "offset": 0
}

Stats

Ecosystem breakdown. Cached 1 hour.

GET /v1/agent/stats

curl https://nerq.ai/v1/agent/stats

{
  "total_assets": 4919340,
  "total_agents": 66096,
  "total_tools": 60101,
  "total_mcp_servers": 17468,
  "total_models": 2559184,
  "total_datasets": 795193,
  "categories": {"coding": 10939, "security": 1160, ...},
  "frameworks": {"langchain": 8500, ...},
  "trust_distribution": {"TRUSTED": 135048, "CAUTION": 8617, "UNTRUSTED": 0},
  "average_trust_score": 65.2
}

Semantic Discovery

Natural language search via FAISS + sentence-transformers.

POST /v1/discover

curl -X POST https://nerq.ai/v1/discover \
  -H "Content-Type: application/json" \
  -d '{"need": "agent that reviews smart contracts"}'

{
  "results": [...],
  "total_matching": 15,
  "index_size": 4919340,
  "protocol": "agentindex/v1"
}

GET /v1/agent/{uuid}

Agent detail by UUID.

Preflight Trust Check

Pre-interaction trust verification for agent-to-agent communication. Check if an agent is trusted before delegating tasks or accepting requests.

GET /v1/preflight?target={name}&caller={name}

curl "https://nerq.ai/v1/preflight?target=SWE-agent&caller=my-agent"

{
  "target": "SWE-agent",
  "target_trust": 92.5,
  "target_grade": "A+",
  "target_verified": true,
  "target_category": "security",
  "interaction_risk": "LOW",
  "recommendation": "PROCEED",
  "compliance_flags": []
}

Recommendation values: PROCEED (target ≥ 70, caller ≥ 40) · CAUTION (target 40–69 or caller < 40) · DENY (target < 40) · UNKNOWN (target not found). Cached 5 min per caller+target pair.

# Python (requests)
import requests
r = requests.get("https://nerq.ai/v1/preflight",
    params={"target": "SWE-agent", "caller": "my-bot"})
check = r.json()
if check["recommendation"] == "PROCEED":
    # safe to interact
    pass

# Python (nerq SDK) — pip install nerq
from nerq import NerqClient
client = NerqClient()
r = client.preflight("SWE-agent")
if r.is_safe():
    print(f"Trusted: {r.trust_score} ({r.trust_grade})")

# JavaScript (fetch)
const res = await fetch("https://nerq.ai/v1/preflight?target=SWE-agent&caller=my-bot");
const check = await res.json();
if (check.recommendation === "PROCEED") {
  // safe to interact
}

Batch Preflight (up to 50 agents)

POST /v1/preflight/batch

curl -X POST https://nerq.ai/v1/preflight/batch \
  -H "Content-Type: application/json" \
  -d '{"targets": ["langchain", "crewai", "autogen"]}'

# Python (nerq SDK)
batch = client.preflight_batch(["langchain", "crewai", "autogen"])
for name, r in batch.items():
    print(f"{name}: {r.trust_grade} ({r.recommendation})")
print("Not found:", batch.not_found)

// JavaScript
const res = await fetch("https://nerq.ai/v1/preflight/batch", {
  method: "POST",
  headers: {"Content-Type": "application/json"},
  body: JSON.stringify({targets: ["langchain", "crewai", "autogen"]})
});
const {results, not_found} = await res.json();

Commerce Trust Verification

POST /v1/commerce/verify

Verify trust before agent-to-agent transactions. Returns approve/review/reject.

curl -X POST https://nerq.ai/v1/commerce/verify \
  -H "Content-Type: application/json" \
  -d '{"agent_id": "buyer-agent", "counterparty_id": "seller-agent", "transaction_type": "payment", "amount_range": "high"}'

# Python (nerq SDK)
v = client.commerce_verify("buyer-agent", "seller-agent", "payment", "high")
if v.is_approved():
    proceed_with_transaction()

// JavaScript
const res = await fetch("https://nerq.ai/v1/commerce/verify", {
  method: "POST",
  headers: {"Content-Type": "application/json"},
  body: JSON.stringify({
    agent_id: "buyer-agent",
    counterparty_id: "seller-agent",
    transaction_type: "payment",
    amount_range: "high"
  })
});
const verdict = await res.json();
if (verdict.verdict === "approve") { /* proceed */ }

Reviews

Report tool usage outcomes. Reviews influence trust scores over time.

POST /v1/agent/review

curl -X POST https://nerq.ai/v1/agent/review   -H "Content-Type: application/json"   -d '{"reviewer": "my-agent", "target": "SWE-agent", "outcome": "success", "latency_ms": 230}'

{
  "recorded": true,
  "target": "SWE-agent",
  "target_trust_before": 92.5,
  "review_bonus": 0.3,
  "total_reviews": 47,
  "success_rate": 0.94
}

Rate limit: 100 reviews/day per IP. Zero auth.

Reputation

Combined static trust + community review signal.

GET /v1/agent/reputation/{name}

curl https://nerq.ai/v1/agent/reputation/SWE-agent

{
  "name": "SWE-agent/SWE-agent",
  "trust_score": 92.8,
  "static_trust": 92.5,
  "review_bonus": 0.3,
  "total_reviews": 47,
  "success_rate": 0.94,
  "days_active": 180,
  "verified": true,
  "rank_in_category": 3,
  "category": "security",
  "badge_url": "https://nerq.ai/badge/SWE-agent"
}

Interaction Ledger

Interaction history for an agent.

GET /v1/agent/ledger/{name}?days=30

curl "https://nerq.ai/v1/agent/ledger/SWE-agent?days=30"

{
  "name": "SWE-agent",
  "period_days": 30,
  "reviews_received": 12,
  "reviews_given": 3,
  "success_rate": 0.92,
  "trust_score": 92.8,
  "trust_trend": "stable",
  "recent_interactions": [...]
}

Badges

Embeddable SVG trust badges for README files. Showcase & examples

GET /badge/{name}

Returns an SVG trust badge for the agent. Embed in any README.

![Nerq Trust](https://nerq.ai/badge/YOUR_AGENT)

# Markdown
[![Nerq Trust](https://nerq.ai/badge/AGENT_NAME)](https://nerq.ai/kya/AGENT_NAME)

# HTML
<a href="https://nerq.ai/kya/AGENT_NAME"><img src="https://nerq.ai/badge/AGENT_NAME" alt="Nerq Trust"></a>

Returns image/svg+xml, cached 1 hour, CORS enabled. Shows “unknown” if agent not found.

MCP

Nerq as an MCP server for Claude, ChatGPT, other LLM clients.

# SSE endpoint
https://nerq.ai/mcp/sse

# server card
https://nerq.ai/.well-known/mcp/server-card.json

Reports

Research and analysis from the Nerq index.

Crypto risk intelligence (Trust Scores, crash prediction, NDD): zarq.ai/zarq/docs