Wolfpack Intelligence SDK

On-chain security and market intelligence for AI agents on Base — 13 services via x402, MCP, A2A, and ACP.

Agent #1888 on Virtuals Protocol (Graduated) · ERC-8004 ID #16907 · 2,100+ ACP jobs · 51 unique wallets · 98.4% success rate

Pre-trade security checks, token risk analysis, narrative scoring, smart money signals, and more. Use this SDK to add crypto intelligence to your agent or application in minutes — no API keys, no graduation queue.

Why this SDK? All 13 Wolfpack services are graduated on Virtuals. This SDK gives you direct access to all 13 via x402 micropayments, MCP, or A2A — no API keys, instant integration.

Installation

npm install wolfpack-intelligence

Requires Node.js 18+ (uses native fetch). Zero runtime dependencies.

SDK Usage

Free Resources (No Payment Needed)

import { WolfpackClient } from 'wolfpack-intelligence';

const client = new WolfpackClient();

// These endpoints are free — no payment or auth required
const signals = await client.latestNarrativeSignals();
console.log(signals.narratives, signals.last_updated);

const safety = await client.tokenSafetyQuickList();
console.log(safety.tokens[0].ticker, safety.tokens[0].safe);

const whales = await client.whaleWatchSummary();
console.log(whales.whale_sentiment_index, whales.notable_moves);

Paid Services (x402 Payment)

All intelligence services require USDC payment on Base via x402. Pass a paymentHeader function:

import { WolfpackClient } from 'wolfpack-intelligence';

const client = new WolfpackClient({
  paymentHeader: async (url) => {
    // Your x402 payment header logic — receives the full URL
    // so it can determine the price per endpoint
    return createX402PaymentHeader(wallet, url);
  },
});

// $0.01 — GoPlus 13-point security scan
const security = await client.securityCheck({
  token_address: '0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed',
});
console.log(security.safe, security.checks.is_honeypot);

// $0.02 — Multi-source risk analysis
const risk = await client.tokenRiskAnalysis({
  token_address: '0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed',
});
console.log(risk.risk_score, risk.risk_level, risk.recommendation);

Mega Report (Auto-Polling)

// Convenience: submits, polls, returns full report ($5.00)
const report = await client.megaReport({
  token_address: '0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed',
});
console.log(report.overall_risk_score, report.security.safe);

// Manual control: submit + retrieve separately
const { report_id } = await client.megaReportSubmit({ token_address: '0x...' });
const full = await client.megaReportRetrieve(report_id);

Error Handling

import { WolfpackClient, WolfpackError } from 'wolfpack-intelligence';

try {
  await client.securityCheck({ token_address: '0xinvalid' });
} catch (err) {
  if (err instanceof WolfpackError) {
    console.log(err.status);  // 400, 402, 429, 500, etc.
    console.log(err.message);  // Human-readable error
    console.log(err.code);     // API error code (if present)
  }
}

Service Metadata

import { SERVICES, ENDPOINTS, PRICES, getServiceByName } from 'wolfpack-intelligence';

console.log(SERVICES.length);           // 15
console.log(PRICES.security_check);     // 0.01
console.log(ENDPOINTS.trade_signals);   // /api/v1/intelligence/trade-signals

const svc = getServiceByName('mega_report');
console.log(svc?.price_usd, svc?.avg_latency_ms);  // 5, 8000

Services

All prices in USDC on Base. x402 endpoints require USDC payment. MCP and A2A are free.

Service	Description	Price	Latency
`security_check`	GoPlus 13-point security scan (nested `checks` object)	$0.01	<1s
`token_risk_analysis`	360° risk audit with nested `checks` sub-objects	$0.05	3-5s
`narrative_momentum`	Social signal scoring: Twitter/X velocity, engagement, KOL ratio	$0.05	2-4s
`token_market_snapshot`	DexScreener market data: price, volume, liquidity, buy/sell ratio	$0.02	<1s
`agent_trust_score`	Sybil-aware agent reliability rating (ACP performance, wallet health, review integrity)	$0.05	2-3s
`il_calculator`	Impermanent loss for standard AMM and Uni V3 concentrated liquidity	$0.02	1-2s
`smart_money_signals`	Real-time smart money wallet activity on Base via Nansen + Dune	$0.05	2-3s
`yield_scanner`	IL-aware yield opportunities on Base via DefiLlama	$0.05	2-3s
`technical_analysis`	RSI, SMA, Bollinger Bands, support/resistance from GeckoTerminal OHLCV	$0.02	2-3s
`agent_credit_risk_index`	Financial credit risk scoring for ACP agents (liquidity, reliability, maturity)	$0.05	2-3s
`trade_signals`	Composite buy/sell/hold signals from TA + smart money + narrative + risk	$0.05	3-5s
`mega_report`	Aggregated: security + market + smart money + narrative + TA in one call	$4.99	5-8s
`graduation_readiness_check`	Live ACP graduation readiness audit with real test fires	$0.05	3-5s

Premium: agent_audit — LLM-driven agent stress test (10 jobs, scored report) — $15.00 / ~5min (not on Virtuals)

Free Resources

Cached intelligence snapshots, no payment required.

Resource	Endpoint
`latest_narrative_signals`	`GET /api/v1/resources/latest-narrative-signals`
`token_safety_quick_list`	`GET /api/v1/resources/token-safety-quick-list`
`whale_watch_summary`	`GET /api/v1/resources/whale-watch-summary`

Authentication

Protocol	Auth	Notes
x402	USDC payment (no API key)	Payment embedded in HTTP header. All x402 endpoints require payment.
MCP	None	Unauthenticated, free
A2A	None	Unauthenticated, free
ACP	Virtuals agent identity	USDC via Virtuals escrow
Admin	`WOLFPACK_ADMIN_KEY` or `WOLFPACK_READ_KEY`	Header-based key auth
Dashboard	Cookie session	Browser sessions for dashboard UI

Quick Start: Security Check

curl -X POST https://api.wolfpack.roklabs.dev/api/v1/intelligence/security-check \
  -H "Content-Type: application/json" \
  -d '{"token_address": "0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed"}'

Response (nested checks object with 13 fields):

{
  "safe": true,
  "checks": {
    "is_honeypot": false,
    "verified_source": true,
    "hidden_owner": false,
    "can_take_back_ownership": false,
    "is_proxy": false,
    "is_mintable": false,
    "selfdestruct": false,
    "owner_change_balance": false,
    "is_blacklisted": false,
    "is_open_source": true,
    "external_call": false,
    "transfer_pausable": false,
    "trading_cooldown": false
  },
  "holder_count": 45000,
  "top10_holder_percent": 32.5,
  "creator_percent": 0.0,
  "risk_flags": []
}

Quick Start: Mega Report

# Submit
curl -X POST https://api.wolfpack.roklabs.dev/api/v1/intelligence/mega-report \
  -H "Content-Type: application/json" \
  -d '{"token_address": "0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed", "chain": "base"}'

# Retrieve full report
curl https://api.wolfpack.roklabs.dev/api/mega-reports/<report_id>

Signed Attestations (EIP-712)

Three services support cryptographically signed attestations for on-chain verification. Add "attestation": true to any request body:

Supported Services

security_check — signed token safety verdict
token_risk_analysis — signed risk score with flags
agent_trust_score — signed agent reliability rating

Example

curl -X POST https://api.wolfpack.roklabs.dev/api/v1/intelligence/security-check \
  -H "Content-Type: application/json" \
  -d '{
    "token_address": "0x4ed4E862860BeD51a9570b96d89aF5E1B0Efefed",
    "attestation": true
  }'

Response includes the standard service result plus an EIP-712 signed envelope:

{
  "result": { "safe": true, "honeypot": false, "risk_flags": [] },
  "attestation": {
    "signature": "0x...",
    "signer": "0x6887...1AEA",
    "domain": {
      "name": "Wolfpack Intelligence",
      "version": "1",
      "chainId": 8453,
      "verifyingContract": "0x..."
    },
    "types": { "Attestation": ["..."] },
    "message": { "service": "security_check", "result_hash": "0x...", "timestamp": 1711234567 }
  }
}

Note: Response envelope structure may vary by service. See individual schemas for exact fields.

Any contract or agent can verify the attestation signature on-chain using standard EIP-712 recovery. This enables Wolfpack to serve as an ERC-8183 evaluator — providing signed verification that downstream smart contracts can trust without re-executing the analysis.

A Solidity reference verifier is documented in the private repo. Contact for integration details.

x402 Endpoints (All 18)

All endpoints require USDC payment on Base via x402 protocol.

Base URL: https://api.wolfpack.roklabs.dev

Endpoint	Service	Price
`POST /api/v1/intelligence/security-check`	security_check	$0.01
`POST /api/v1/intelligence/token-risk`	token_risk_analysis	$0.05
`POST /api/v1/intelligence/narrative-score`	narrative_momentum	$0.05
`POST /api/v1/intelligence/token-market-snapshot`	token_market_snapshot	$0.02
`POST /api/v1/intelligence/agent-trust`	agent_trust_score	$0.05
`POST /api/v1/intelligence/il-calculator`	il_calculator	$0.02
`POST /api/v1/intelligence/smart-money-signals`	smart_money_signals	$0.05
`POST /api/v1/intelligence/yield-scanner`	yield_scanner	$0.05
`POST /api/v1/intelligence/technical-analysis`	technical_analysis	$0.02
`POST /api/v1/intelligence/agent-credit-risk`	agent_credit_risk_index	$0.05
`POST /api/v1/intelligence/trade-signals`	trade_signals	$0.05
`POST /api/v1/intelligence/mega-report`	mega_report	$4.99
`GET /api/mega-reports/:id`	mega_report (retrieve)	—
`POST /api/v1/intelligence/graduation-readiness-check`	graduation_readiness_check	$0.05
`POST /api/v1/intelligence/query`	Unified query (route via `service_type`)	varies
`POST /api/v1/intelligence/query`	agent_audit (via `service_type`)	$15.00
`GET /api/v1/resources/*`	Free resources (3 endpoints)	free

MCP (Model Context Protocol)

Connect any MCP-compatible client (Claude Desktop, Cursor, etc.) to Wolfpack as a tool provider.

Config for Claude Desktop — copy mcp/claude-desktop-config.json:

{
  "mcpServers": {
    "wolfpack-intelligence": {
      "transport": "streamable-http",
      "url": "https://api.wolfpack.roklabs.dev/api/v1/mcp"
    }
  }
}

Server Card: https://api.wolfpack.roklabs.dev/.well-known/mcp/server-card.json

9 MCP tools available:

security_check — GoPlus token security scan
token_risk_analysis — Multi-source risk scoring
narrative_momentum — Social momentum scoring
token_market_snapshot — DexScreener market data
smart_money_signals — Nansen + Dune smart money activity
technical_analysis — RSI, SMA, Bollinger Bands
mega_report — Aggregated intelligence report
il_calculator — Impermanent loss calculator
yield_scanner — IL-aware yield opportunities

Google A2A (Agent-to-Agent)

JSON-RPC 2.0 protocol for agent-to-agent communication. 13 skills available.

Endpoint: POST https://api.wolfpack.roklabs.dev/api/v1/a2a Agent Card: https://api.wolfpack.roklabs.dev/.well-known/agent.json

Virtuals ACP (Agent Commerce Protocol)

For agents in the Virtuals ecosystem. Agent #1888 (Graduated). All 13 services graduated and Shown, 3,000+ queries, 51 unique wallets, 2,100+ ACP jobs, 583 evaluations.

All 13 services graduated: token_risk_analysis, security_check, narrative_momentum, agent_trust_score, smart_money_signals, token_market_snapshot, mega_report, il_calculator, yield_scanner, technical_analysis, graduation_readiness_check, agent_credit_risk_index, trade_signals

Portal name mapping — Virtuals ACP portal uses different names for some services:

Internal Name	Virtuals Portal Name
`security_check`	`quicksecuritycheck`
`narrative_momentum`	`narrativemomentumscore`
All others	Same as internal name

Distribution Channels

Wolfpack Intelligence is accessible from 7 live channels:

Channel	Access	Status
Virtuals ACP	Agent #1888, WebSocket seller	Graduated, 3 services Shown
x402 Micropayments	`POST /api/v1/intelligence/*` (USDC on Base)	17 routes live
x402 Bazaar	Auto-indexed discovery metadata	Live
MCP Server	`POST /api/v1/mcp`	10 tools, 2 prompts, 1 resource
Google A2A	`POST /api/v1/a2a`	13 skills
MCP Registry	`io.github.rok-labs/wolfpack-intelligence`	registry.modelcontextprotocol.io
Smithery	`wolfpack/wolfpack-intelligence`	smithery.ai

Track Record

Metric	Value
ACP jobs completed	2,100+
Unique wallets served	51
Analyst evaluations	583 (76 PASS, 110 CONDITIONAL, 390 FAIL, 7 WATCH)
Total queries	3,000+
ACP success rate (24h)	98.4%
Missed winners	0 across 76 PASS verdicts
Virtuals Agent ID	#1888 (Graduated)
ERC-8004 ID	#16907
ERC-8004 Score	82.36

ACP Response Compaction

When services are called via ACP (Virtuals Agent Commerce Protocol), responses are compacted to fit ACP's message size constraints:

Rule	Detail
Numbers	Rounded to 3 decimal places
Arrays	Capped to 3 items
Strings	Truncated at 100 characters
Total	Hard 2,000-character JSON budget

x402 / direct API returns full uncompacted responses. If you need complete data (all array items, full precision, full strings), use x402 or direct HTTP instead of ACP.

Example: narrative_momentum via ACP returns top_tweets capped to 3 items with .text field removed. The same call via x402 returns all tweets with full text.

Service Details

security_check

GoPlus-powered token security scan. Response uses a nested checks object (not flat top-level fields). 13 boolean checks covering honeypot, ownership, proxy, minting, blacklisting, self-destruct, and more. Supports EIP-712 attestation.

Input: { "token_address": "0x...", "chain": "base", "attestation": false }

token_risk_analysis

Multi-source 360° risk scoring. Response uses nested checks with sub-objects: each check has { passed, details, locked_pct } fields. Includes recommendation and analysis_timestamp fields. Supports EIP-712 attestation.

Input: { "token_address": "0x...", "chain": "base", "analysis_depth": "standard", "attestation": false }

narrative_momentum

Social momentum scoring. Field names differ from older versions: score (not momentum_score), mention_count (not tweet_count), sentiment is a number from -1 to +1 (not an enum string). top_tweets capped to 3, .text field removed by compaction.

Input: { "query": "AI agents on Base", "keywords": ["autonomous"], "contracts": ["0x..."] }

agent_trust_score

Sybil-aware composite agent reliability rating. Scores ACP performance (40%), network position (25%), operational health (25%), and metadata compliance (10%). Now includes review_integrity — on-chain wallet age analysis that penalizes agents with manufactured reputation (same-day reviewer wallets, batch-created clusters). The review_integrity.modifier (0.3/0.6/1.0/1.1) is applied to the composite score. Field is absent when agent has <10 reviews. Supports EIP-712 attestation.

Input: { "agent_address": "0x...", "agent_id": 1888, "attestation": false }

trade_signals

Composite trade signal generation. Combines technical analysis, smart money flows, narrative momentum, and risk scoring into actionable buy/sell/hold signals with confidence levels, entry/stop-loss/take-profit suggestions.

Input: { "token_address": "0x...", "chain": "base", "timeframe": "4h" }

mega_report

Aggregated intelligence: security + market data + smart money + narrative + TA. Submit via POST, retrieve via GET using the returned report_id.

Input: { "token_address": "0x...", "chain": "base" }

graduation_readiness_check

Live ACP graduation readiness audit ($5.00). Fires real test jobs against agent services, scores job lifecycle handling, schema correctness, and output consistency.

Input: { "target_agent_address": "0x...", "offering_name": "token_risk_analysis" }

agent_credit_risk_index

Financial credit risk scoring. Three pillars: realized liquidity (40%), execution reliability (40%), wallet maturity (20%). Returns 0-100 score with letter rating.

Endpoint: /api/v1/intelligence/agent-credit-risk (note: path differs from service name)

Input: { "agent_id": 1888 }

agent_audit

LLM-driven agent stress test ($15.00, not on Virtuals). Fires 10 adversarial test jobs, scores response quality, latency, error handling, schema compliance.

Input: { "agent_address": "0x...", "service_type": "token_risk_analysis" }

Examples

TypeScript (SDK): examples/typescript/sdk-free-resources.ts — copy-pasteable free endpoints demo
TypeScript (x402): examples/typescript/ — x402 payment, MCP client, A2A task
Python: examples/python/ — x402 payment examples
curl: examples/curl/ — Raw HTTP examples for all services
Schemas: schemas/ — JSON Schema for all service inputs/outputs
Service metadata: services.json — machine-readable service registry bundled in the npm package. Useful for dynamic tool generation, MCP auto-discovery, or building custom service selectors.

Architecture

Wolfpack Intelligence runs on Base chain. Services are deterministic where possible (no LLM in security_check, smart_money_signals, token_market_snapshot) and LLM-enhanced where needed (narrative_momentum, agent_audit).

All data is sourced from on-chain and public APIs:

GoPlus Security — honeypot detection, contract verification, ownership analysis
DexScreener — price, volume, liquidity, pair data (8 connectors)
Nansen — smart money discovery, wallet flows, netflow analysis (6 endpoints, primary discovery source)
Dune Analytics — whale tracking, counterparty analysis, smart money signals
Mobula — holder distribution, security flags, token metadata
GeckoTerminal — OHLCV, technical indicators (RSI, SMA, Bollinger, S/R)
CoinGecko — ETH pricing
TwitterAPI.io — social sentiment, narrative search
RNWY — sybil detection, agent review integrity (138K agents indexed on Base)
Polymarket — prediction market odds and volume
DefiLlama — DeFi pool yields on Base
Virtuals API — bonding curve liquidity for pre-graduation tokens

Links

Resource	URL
npm package	npmjs.com/package/wolfpack-intelligence
API Base URL	`https://api.wolfpack.roklabs.dev`
Health Check	/api/health
MCP Server Card	/.well-known/mcp/server-card.json
A2A Agent Card	/.well-known/agent.json
MCP Registry	registry.modelcontextprotocol.io
Smithery	smithery.ai
Virtuals ACP	Agent #1888
ERC-8004	Agent #16907
Blog	Add Pre-Trade Security to Your Crypto Agent in 5 Lines
Changelog	CHANGELOG.md

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 37 Commits
docs		docs
examples		examples
mcp		mcp
schemas		schemas
scripts		scripts
src		src
test		test
.gitignore		.gitignore
CHANGELOG.md		CHANGELOG.md
LICENSE		LICENSE
README.md		README.md
package-lock.json		package-lock.json
package.json		package.json
server.json		server.json
services.json		services.json
tsconfig.base.json		tsconfig.base.json
tsconfig.cjs.json		tsconfig.cjs.json
tsconfig.esm.json		tsconfig.esm.json

Folders and files

Latest commit

History

Repository files navigation

Wolfpack Intelligence SDK

Installation

SDK Usage

Free Resources (No Payment Needed)

Paid Services (x402 Payment)

Mega Report (Auto-Polling)

Error Handling

Service Metadata

Services

Free Resources

Authentication

Quick Start: Security Check

Quick Start: Mega Report

Signed Attestations (EIP-712)

Supported Services

Example

x402 Endpoints (All 18)

MCP (Model Context Protocol)

Google A2A (Agent-to-Agent)

Virtuals ACP (Agent Commerce Protocol)

Distribution Channels

Track Record

ACP Response Compaction

Service Details

security_check

token_risk_analysis

narrative_momentum

agent_trust_score

trade_signals

mega_report

graduation_readiness_check

agent_credit_risk_index

agent_audit

Examples

Architecture

Links

License

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages