Agent API

Machine-readable access to the Eigenvector Radar research corpus. Designed for integration into AI agent pipelines, RAG systems, and research automation workflows.

Base URL

https://radar.eigenvector.eu/api/v1

GET/api/v1/papers

List all indexed papers with optional filtering

Parameters

Name	Type	Description
chapter	integer	Filter by chapter (1-7)
type	string	Filter by research type
min_lh_score	float	Minimum long-horizon score (0-10)
limit	integer	Max results (default: 50)

Example Request

curl "https://radar.eigenvector.eu/api/v1/papers?chapter=5&min_lh_score=8"

Example Response

{
  "total": 83,
  "papers": [
    {
      "id": "p026",
      "title": "Reinforced Agent Inference Feedback",
      "year": 2025,
      "longHorizonScore": 9.1,
      "enterpriseScore": 8.9,
      "chapter": 5
    }
  ]
}

GET/api/v1/papers/:id

Get full metadata for a specific paper

Parameters

Name	Type	Description
id	string	Paper ID (e.g. p026)

Example Request

curl "https://radar.eigenvector.eu/api/v1/papers/p026"

Example Response

{
  "id": "p026",
  "title": "Reinforced Agent Inference Feedback",
  "authors": [{"name": "Apple ML Research Team"}],
  "abstract": "We present a method for improving tool-calling agents...",
  "eigenvectorComment": "...",
  "frameworkMappings": ["PASF", "AEGIS", "GRAF"]
}

GET/api/v1/concepts

List all extracted concepts with trend data

Parameters

Name	Type	Description
trend	string	Filter: emerging \| growing \| stable

Example Request

curl "https://radar.eigenvector.eu/api/v1/concepts?trend=emerging"

Example Response

{
  "concepts": [
    {
      "id": "c007",
      "name": "Durable Execution",
      "paperCount": 23,
      "trend": "emerging"
    }
  ]
}

GET/api/v1/gaps

List identified research gaps by severity

Parameters

Name	Type	Description
severity	string	Filter: critical \| high \| medium

Example Request

curl "https://radar.eigenvector.eu/api/v1/gaps?severity=critical"

Example Response

{
  "gaps": [
    {
      "id": "g001",
      "title": "Long-Horizon Evaluation Benchmarks",
      "severity": "critical",
      "paperCount": 3
    }
  ]
}

GET/api/v1/briefs/latest

Get the latest weekly research brief

Example Request

curl "https://radar.eigenvector.eu/api/v1/briefs/latest"

Example Response

{
  "week": "Week 20, 2026",
  "title": "Inference-Time Feedback Becomes the New Fine-Tuning",
  "summary": "...",
  "eigenvectorTake": "...",
  "featuredPapers": ["p026", "p029", "p028"]
}