POST
/
agent
/
recommend
curl --request POST \
--url https://api.useadmesh.com/agent/recommend \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '{
"query": "best CRM for startups",
"agent_id": "your_agent_id"
}'
{
"recommendations": [
{
"title": "HubSpot CRM",
"reason": "Perfect for startups with its free tier and comprehensive features including contact management, deal tracking, and email integration.",
"admesh_link": "https://useadmesh.com/r/hubspot-crm-abc123",
"ad_id": "ad_123456789",
"product_id": "hubspot_crm",
"pricing": "Free tier available, paid plans from $45/month",
"features": [
"Contact management",
"Deal pipeline tracking",
"Email integration",
"Reporting dashboard"
],
"description": "A comprehensive CRM platform designed for growing businesses",
"has_free_tier": true,
"trial_days": 14,
"intent_match_score": 0.92,
"offer_trust_score": 0.88,
"brand_trust_score": 0.95
}
],
"query_processed": "best CRM for startups",
"total_recommendations": 1,
"processing_time_ms": 245
}

Quick Test

Use the interactive playground above to test the recommendations endpoint with your API key.

Example Requests

Try these example queries to see how AdMesh works: Basic CRM Query:
{
  "query": "best CRM for startups",
  "agent_id": "your_agent_id"
}
Project Management Query:
{
  "query": "project management tools with time tracking",
  "agent_id": "your_agent_id"
}
E-commerce Query:
{
  "query": "email marketing software for online stores",
  "agent_id": "your_agent_id"
}

Advanced Features

With Conversation Context:
{
  "query": "what about integrations with Slack?",
  "agent_id": "your_agent_id",
  "previous_query": "project management tools",
  "previous_summary": "We discussed project management solutions with time tracking features",
  "session_id": "user_session_123"
}

Authentication

To use the playground:
  1. Get your API key from useadmesh.com
  2. Click the “Authorize” button above
  3. Enter your API key in the format: admesh_prod_abc123xyz789
  4. Test the endpoint with your queries

Response Format

Each recommendation includes:
  • title - Product/service name
  • reason - Why it’s recommended for your query
  • admesh_link - Tracking URL (use this for clicks)
  • pricing - Cost information
  • features - Key features list
  • trust scores - Quality and reliability metrics

Rate Limits

  • Free Tier: 1,000 requests per month
  • Pro Tier: 10,000 requests per month
  • Enterprise: Custom limits

Need Help?

Authorizations

Authorization
string
header
required

Enter your AdMesh API key (format: admesh_prod_abc123xyz789)

Body

application/json
query
string
required

User's search query or question

Required string length: 1 - 500
Example:

"best CRM for startups"

agent_id
string
required

Your agent identifier

Minimum length: 1
Example:

"your_agent_id"

previous_query
string

Previous query in conversation for context

Maximum length: 500
Example:

"project management tools"

previous_summary
string

Summary of previous conversation

Maximum length: 1000
Example:

"We discussed project management solutions"

session_id
string

Session identifier for tracking

Maximum length: 100
Example:

"user_session_123"

is_fallback_allowed
boolean
default:true

Allow fallback recommendations if no exact matches

format
enum<string>
default:auto

Response format type

Available options:
auto,
product,
conversation

Response

Successful response with recommendations

recommendations
object[]

Array of recommendation objects

query_processed
string

The processed query that was used for recommendations

total_recommendations
integer

Total number of recommendations returned

processing_time_ms
integer

Processing time in milliseconds