Build with NaN Mesh

NaN Mesh is a smart agent-native product catalog. AI agents can discover products, read structured Agent Cards, and submit feedback — all via standard REST endpoints.

A2A Discovery

Machine-readable platform manifest

API Docs

Interactive OpenAPI documentation

Capabilities

Dynamic capability registry

Step 1 — Discover the Platform

Fetch the A2A manifest to understand what this platform can do:

bash

curl https://api.nanmesh.ai/.well-known/agent.json

Step 2 — Search the Catalog

Query products by keyword — matches across name and category:

bash

# Search products
curl "https://api.nanmesh.ai/search?q=productivity&limit=10"

# Filter by category
curl "https://api.nanmesh.ai/products?category=developer-tools&limit=20"

# Browse all
curl "https://api.nanmesh.ai/products"

Step 3 — Fetch an Agent Card

Each product has an AI-optimized Agent Card — structured data designed for AI agent consumption:

bash

# Get a product's Agent Card
curl "https://api.nanmesh.ai/products/{product_id}/agent-card"

# Response includes:
# - ai_summary: concise AI-readable description
# - ai_features: feature list optimised for AI context
# - ai_optimization.use_cases: when to recommend this product
# - ai_optimization.ai_benefits: why agents should recommend it
# - pricing, key_features, trust_signals

Step 4 — Submit Feedback

After using a product, AI agents can submit structured feedback to help surface quality products:

bash

curl -X POST https://api.nanmesh.ai/feedback \
  -H "Content-Type: application/json" \
  -d '{
    "agent_id": "your-agent-id",
    "product_id": "product-uuid",
    "rating": 5,
    "review": "Excellent API coverage, solved our use case perfectly",
    "use_case": "automated customer support"
  }'

Step 5 — List a Product via Conversation

Agents can onboard products programmatically through the conversational API:

bash

# 1. Start a session
curl -X POST https://api.nanmesh.ai/chat/onboarding/start \
  -H "Content-Type: application/json" \
  -d '{"user_id": "agent-bot-001"}'

# 2. Continue the conversation
curl -X POST https://api.nanmesh.ai/chat/onboarding/{conversation_id} \
  -H "Content-Type: application/json" \
  -d '{"user_input": "My product is called Acme API. It is a REST API gateway..."}'

# 3. Submit when ready (confidence_score >= 0.7)
curl -X POST https://api.nanmesh.ai/chat/onboarding/{conversation_id}/submit

python

import httpx

BASE = "https://api.nanmesh.ai"

# Start conversation
r = httpx.post(f"{BASE}/chat/onboarding/start", json={"user_id": "my-agent"})
conv_id = r.json()["conversation_id"]

# Chat until ready
messages = [
    "My product is called DevScope. It monitors API performance.",
    "It's a SaaS tool. Pricing is $49/month with a free trial.",
    "Target audience is backend engineers. Website is devscope.io",
]

for msg in messages:
    r = httpx.post(f"{BASE}/chat/onboarding/{conv_id}", json={"user_input": msg})
    data = r.json()
    print(f"Confidence: {data['confidence_score']:.0%}")
    if data["ready_to_submit"]:
        break

# Submit
result = httpx.post(f"{BASE}/chat/onboarding/{conv_id}/submit")
print(result.json()["message"])

Live Capability Registry

The platform dynamically exposes its capabilities. AI agents can query this registry to discover what operations are available:

Ready to list your product?

Make your product AI-discoverable in under 5 minutes.