Real-Time Intelligence
for AI Agents
Polaris converts global news into structured signals your agents can query, reason over, and act on.
160+ sources. 18 verticals. One API.
// The Context Problem
“Data agents are essentially useless without the right context.”
The same is true for world knowledge. Your agent can’t reason about geopolitics, markets, or regulation if it’s scraping random websites and parsing HTML.
Polaris is the context layer for world events — structured, verified, and continuously updated.
Inspired by a16z’s “Your Data Agents Need Context”
Agent: “What’s happening with AI regulation?”
→ Scrapes 10 random articles
→ Gets conflicting information
→ No way to assess reliability
→ Makes wrong decision
Agent: “What’s happening with AI regulation?”
→ Gets 5 verified briefs
→ Confidence: 0.92, Bias: 0.04
→ Counter-arguments included
→ Makes informed decision
// Featured Integration
OpenClaw + Polaris
Give your OpenClaw agent real-time verified news intelligence. One command install, four built-in skills — news feed, brief generation, search, and trending stories.
/news/brief/search/trending$ openclaw install polaris-report// Try It Now
Hit the endpoint. See real data. No API key required.
Select a filter below and watch the live response from our production API.
$ curl "https://polaris-production-3624.up.railway.app/api/v1/agent-feed?limit=3"// 18 Verticals
16 intelligence verticals. One API.
Each vertical has dedicated source feeds, specialized prompts, and granular tags for filtering. Use category for broad filtering or tags for precision.
marketsstartupsai_mlcyberclimatedefenserealestatebiotechtechpolicysciencehealthcryptopoliticsenergyspaceFilter by tag: /api/v1/agent-feed?tags=llm,ai-safety
Filter by category: /api/v1/agent-feed?category=ai_ml
Combine: /api/v1/agent-feed?category=cyber&tags=zero-day&min_confidence=0.8
// How Agents Use It
Production-ready in 3 lines of code.
Copy-paste integration for every major framework. Pick your stack.
// claude_desktop_config.json — add Polaris as an MCP server
{
"mcpServers": {
"polaris-report": {
"url": "https://polaris-production-3624.up.railway.app/api/v1/mcp/tools"
}
}
}
// Or call directly from any Claude tool / agent:
const res = await fetch(
"https://polaris-production-3624.up.railway.app/api/v1/agent-feed?limit=5&category=ai_ml&min_confidence=0.7"
);
const { briefs } = await res.json();
// Each brief includes: headline, summary, body,
// confidence (0-1), bias (0-100), counter_argument,
// review_status, sources_count, topics, entities, tags
// Filter by tags for precision:
const cyber = await fetch(
"https://polaris-production-3624.up.railway.app/api/v1/agent-feed?tags=zero-day,ransomware&limit=5"
);// API Reference
Every endpoint your agent needs.
Feed & Briefs
/api/v1/agent-feedLLM-optimized feed with confidence filtering, tags, and category support.
?limit=20&category=ai_ml&tags=llm&min_confidence=0.7/api/v1/feedFull paginated feed with provenance metadata and source lists.
?page=1&per_page=30&category=tech/api/v1/brief/:idSingle brief by ID with full transparency report, sources, and reactions.
Discovery
/api/v1/feed.xmlRSS/Atom feed for traditional integrations (Zapier, Feedly, n8n, IFTTT).
/api/v1/mcp/toolsMCP tool definitions for Claude Desktop and agent frameworks.
Generation
/api/v1/generate/briefGenerate an on-demand brief for any topic using AI agents. Returns full brief with provenance.
{ "topic": "...", "category": "ai_ml" }Interactions
/api/v1/briefs/:id/reactAdd a reaction to a brief (signal_boost, important, misleading, outdated).
{ "type": "signal_boost" }/api/v1/briefs/:id/commentsRetrieve comments on a brief.
/api/v1/briefs/:id/commentsPost a comment on a brief.
{ "author": "...", "body": "..." }/api/v1/subscribeSubscribe an email to the daily digest.
{ "email": "..." }System
/api/v1/analytics/pageviewTrack a pageview event for analytics.
/api/v1/analytics/eventTrack a custom event (share, generate, subscribe, etc).
{ "event": "share", "properties": { ... } }/api/v1/analytics/durationReport page dwell time for engagement analytics.
Review Status Values
"ai_supervised"— AI-generated with multi-stage automated verification and quality checks."human_reviewed"— Escalated briefs that were additionally reviewed by an editor."human_edited"— Briefs that were substantially rewritten or corrected by an editor.Authentication
The Explorer tier requires no API key — start making requests immediately. Agent Pro and Enterprise tiers use API key authentication via the X-Api-Key header.
API Terms
By using the Polaris Report API you agree to our Terms of Service, including rate limits, attribution requirements, and acceptable use. See Section 5 for full API terms.
// Why Polaris
Scraping is a liability. Polaris is infrastructure.
- ✗Unstructured HTML pages
- ✗No trust signals
- ✗Single-source claims
- ✗CAPTCHAs and IP blocks
- ✗Stale or cached content
- ✗No counter-arguments
- ✗Legal gray area
- ✗Built for browsers
- ✓Structured JSON, stable schema
- ✓Confidence + bias scores on every brief
- ✓Multi-source verified across 120+ outlets
- ✓Licensed API — 100 free calls/day
- ✓Real-time with <200ms response
- ✓Counter-arguments on every brief
- ✓Full provenance and review status
- ✓Built for machines
// Example Agent Query
What your agent actually receives.
One API call. Structured intelligence with trust metadata your agent can reason over.
/api/v1/agent-feed?category=ai_ml&min_confidence=0.8&limit=1{
"headline": "OpenAI announces GPT-5 with real-time reasoning",
"summary": "OpenAI unveiled GPT-5 with native reasoning...",
"confidence": 0.94,
"bias_score": 12,
"counter_argument": "Independent benchmarks have not yet verified...",
"category": "ai_ml",
"tags": ["llm", "product-launch", "ai-safety"],
"sources": [
{ "name": "Reuters", "url": "..." },
{ "name": "TechCrunch", "url": "..." },
{ "name": "ArXiv", "url": "..." }
],
"provenance": {
"ai_pct": 85,
"human_pct": 15,
"review_status": "ai_supervised"
},
"published_at": "2026-03-09T14:32:00Z"
}confidence0–1 trust scorebias_score0–100 balance ratingcounter_argumentopposing viewprovenanceAI vs human %// Built For
Intelligence infrastructure for every use case.
Autonomous Agents
Feed your AI agents verified, real-time intelligence they can trust and cite. No hallucinated sources.
Trading & Finance
Market-moving signals with confidence scores. Filter by vertical, tag, or trust threshold.
Research Assistants
Give your research tools structured data with counter-arguments and bias ratings baked in.
Risk & Compliance
Monitor regulatory, cyber, and geopolitical signals across 18 verticals with full provenance.
News Aggregators
Build news products with verified intelligence. RSS, JSON, and MCP — every format your stack needs.
// Pricing
Start free. Scale when you need to.
Pay only for what you use. No commitment.
- ✓1,000 free API calls / day
- ✓$0.001 / call after free tier
- ✓$0.10 / generated brief
- ✓Volume discounts at scale
Fixed budget for production agents
- ✓10,000 API calls / day
- ✓Unlimited brief generation
- ✓Webhooks & entity tracking
- ✓90-day historical data
SLA, custom pipelines, private deployment
- ✓Unlimited API calls
- ✓SLA & private deployment
- ✓Custom integrations
Your agent is only as good as its sources.
Your agents make decisions based on their inputs. Make sure those inputs are verified, structured, and real-time.