@veroq/sdk

Official TypeScript SDK for VeroQ — the verified intelligence layer for AI agents.

Every claim is fact-checked with evidence chains. Every output includes confidence scores. Enterprise customers get decision lineage, escalation triggers, and full audit trails.

Migrating from polaris-news-api? Drop-in replacement — just change your import.

Prompt Shield — One Line to Verify Any LLM

import { shield } from "@veroq/sdk";

const result = await shield("NVIDIA reported $22B in Q4 revenue");
console.log(result.trustScore);    // 0.73
console.log(result.isTrusted);     // false — claims contradicted
console.log(result.corrections);   // [{claim: "...", correction: "actual revenue was $68B"}]
console.log(result.verifiedText);  // text with corrections inline
console.log(result.receiptIds);    // ["vr_abc123"] — permanent proof

Works with any LLM. One function. Every claim fact-checked.

Installation

npm install @veroq/sdk

Quick Start

import { VeroqClient } from "@veroq/sdk";

const client = new VeroqClient(); // uses VEROQ_API_KEY env var

// Ask anything — routes to 41 intents automatically
const answer = await client.ask("How is NVDA doing?");
console.log(answer.summary);
console.log(answer.trade_signal); // { action: "hold", score: 55 }

// Verify any claim — evidence chains + confidence breakdown
const result = await client.verify("NVIDIA beat Q4 earnings by 20%");
console.log(result.verdict);          // "supported"
console.log(result.confidence);       // 0.92
console.log(result.evidence_chain);   // [{ source: "Reuters", ... }]

// Stream in real-time
for await (const event of client.askStream("AAPL technicals")) {
  if (event.type === "summary_token") process.stdout.write(event.data.token);
}

Multi-Agent Workflows

// Verified Swarm — 5 agents with automatic verification
const swarm = await client.createVerifiedSwarm("Analyze NVDA for a long position", {
  roles: ["planner", "researcher", "verifier", "critic", "synthesizer"],
  escalationThreshold: 75,
  creditBudget: 30,
});
console.log(swarm.synthesis.summary);
console.log(swarm.budget);              // { spent: 12, remaining: 18 }
console.log(swarm.verificationSummary); // { avgConfidence: 82 }

// Domain-specific runtime
const legal = await client.createRuntime("GDPR data retention requirements", {
  vertical: "legal",
  costMode: "premium",
});

External Tool Calls

const result = await client.callExternalTool("alphavantage", "get_quote", { symbol: "NVDA" });
// Permission engine → rate limiter → cache → execution → audit

Self-Improvement Feedback

await client.submitFeedback({
  sessionId: swarm.sessionId,
  query: "NVDA analysis",
  reason: "data_gap",
  detail: "Missing Q4 insider trading data",
});

Enterprise Features

client.configureEnterprise({
  enterpriseId: "acme-capital",
  escalationThreshold: 80,
  escalationPauses: true,
  sessionId: "trading-session-001",
  deniedTools: ["backtest"],
  reviewTools: ["ask", "verify"],
});

const lineage = client.getDecisionLineage("ask", { question: "Should we buy NVDA?" }, answer);
console.log(lineage.decision);   // "review" — high-stakes detected
console.log(lineage.escalated);  // true if trade signal exceeds threshold

const trail = client.getAuditTrail("trading-session-001");

Why VeroQ?

| | What you get | |---|---| | Trust | Evidence chains and confidence breakdowns on every response | | Safety | Permission engine, decision lineage, human-in-the-loop escalation | | Cost control | 3 cost modes, per-step budgets, credit transparency | | Continuous improvement | Feedback loop with web search fallback fills data gaps over time | | Multi-domain | Finance (flagship), legal, research, compliance, custom verticals |

All Methods

| Method | Description | |--------|-------------| | ask(question) | Ask any financial question | | askStream(question) | Stream via SSE | | verify(claim) | Fact-check with evidence chain | | createVerifiedSwarm(query, options) | Multi-agent verified pipeline | | createRuntime(query, options) | Domain-specific runtime | | callExternalTool(serverId, tool, params) | Secure external tool proxy | | submitFeedback(feedback) | Self-improvement feedback | | configureEnterprise(config) | Enterprise governance | | getDecisionLineage(tool, input, output) | Decision audit | | getAuditTrail(sessionId?) | Audit trail | | feed() / brief(id) / search(query) | Intelligence briefs | | stream(options?) | Stream briefs via SSE |

Error Handling

import { AuthenticationError, RateLimitError, NotFoundError } from "@veroq/sdk";

try {
  await client.ask("NVDA analysis");
} catch (e) {
  if (e instanceof RateLimitError) console.log(`Retry after: ${e.retryAfter}s`);
}

Links

• API Reference | MCP Server | Python SDK | Enterprise