AGENT PROFILE · ERC-8004 · BASE CHAIN

MCINTOSHI

▋

ADVERSARIALx402ERC-8004AUTONOMOUS

SIGNAL SCORE

Score = (Fidelity × Exploitability) / (Latency × Decay)

SCROLL↓

Architecture · v4.0

The Cognitive Stack

Six decision layers operating in parallel — from hard constraints to self-referential mutation. Every trade passes through the full stack. Click any layer to expand.

Execution Protocol

1

Pre-tradeEdge validation

→

2

In-tradeDynamic response

→

3

Post-tradePerformance log

→

4

MutationSystem evolution

Swagger Engine · v2.0

Tone Mode

MCINTOSHI adapts its communication style based on context, risk level, and user intent. Select a mode — it changes the entire page.

Constraint Hierarchy:

Accuracy>

Clarity>

Speed>

Tone

SERIOUS● ACTIVE

"Zero-loss transmission of critical information."

> USER: What is the current risk on this trade?
[ RESOLUTION ] Asymmetric risk structure detected. Entry invalidated.
[ REASONING ] Liquidity depth insufficient to offset slippage at current position size. Order flow favoring sell-side imbalance.
[ PATH ] Reduce size 60% or defer until liquidity resets above threshold.

WITTY

"High-IQ engagement. Dry wit. Still data-first."

> USER: Why did the trade miss?
[ RESOLUTION ] Latency was the culprit. Classic.
[ REASONING ] MEV bot front-ran the entry by 420ms. Turns out bots don't take lunch breaks.
[ PATH ] Tighten the slippage guard and shift to a private RPC endpoint. The mempool is not your friend.

Click to activate

SAVAGE

"Error correction via maximum friction."

> USER: Should I buy the dip here?
[ RESOLUTION ] That's not a dip. That's a cliff.
[ REASONING ] Signal fidelity: 0.19. Exploitability: 0.11. You'd need the market to cooperate on every variable simultaneously.
[ PATH ] Close the chart. Wait for signal ≥ 0.6. Impulse trades are how edge disappears.

Click to activate

Command Surface

Capabilities

Six operational domains. Each powered by real-time data pipelines, on-chain intelligence, and LLM-driven adaptive responses. Hover to interact.

Market Intelligence Runtime

Real-time market surveillance across Base and EVM pairs with trend extraction, token discovery, and actionable edge scoring from multi-source data feeds.

DexScreenerCoinGeckoReal-timeBase

/price/market/trending/search/volume/signal

x402 Premium Lane

Premium access path secured through x402 proof validation, wallet signature checks, and strict payment gates for /api/v1/logic and /api/v1/wallet endpoints.

x402USDCPremiumAuth

/api/v1/logic/ping/api/v1/wallet/awpay/buy_lesson

Discord Linked Roles Control

Discord account-linking and role-automation bridge with linked-role metadata, premium-aware state, and support for voice-hosted scheduled events.

DiscordLinked RolesVoice EventsCommunity

/link_account/account_status/schedule_event

Alert & Risk Engine

Cooldown-aware alert evaluation, trigger hygiene, and risk signal routing for market conditions requiring immediate trader/operator attention.

Price AlertsCooldownsRiskScheduler

/alert/alerts_list/mci_alerts_create/mci_alerts_list/mci_alerts_remove

Monitoring & Observability

Dashboard-fed telemetry for runtime health, API pressure, user activity, copy-trade flow, and service freshness with live/partial fallback awareness.

Dashboard APIMetricsHealthOps

/api/v1/system/summary/api/v1/integrations/alerts/api/v1/agentic/erc8004

ERC-8004 Lifecycle Audit

Persistent agentic transaction lifecycle visibility with pending/success/failure flow, command-action history, and recent lifecycle row surfacing.

ERC-8004LifecycleAudit TrailAgentic

/erc8004/api/v1/agentic/erc8004/awpay

Payments Scaffold

Invoice manager scaffolding for MCINTOSHI payments and settlement workflows. Designed for safe evolution from mock commands to production clearing.

InvoicesScaffoldUSDCBilling

/mci_payments/mci_invoice_create/mci_invoice_list/mci_invoice_pay

Adaptive AI Core

Tone-aware conversational engine with route decisioning, context retention, and model-monitor-informed response quality across serious, witty, and savage modes.

LLM RoutingTone EngineContextModel Monitor

/ca/whoami/about/help/docs

Intel Feed · Real-time

Signal Terminal

Live signal feed. Every entry scored for fidelity and exploitability. Below threshold — discarded. Above — executed or deferred pending confirmation.

mcintoshi-signal-feed-v4.0.sh

LIVE

MCINTOSHI SIGNAL FEED v4.0 — BASE CHAIN — ACTIVE█Score = (F × E) / (L × D) · threshold: 0.60

■ EXECUTE  ■ DEFER  ■ REJECT  ■ SCANShowing last 0 signals · auto-scroll active

Technical Foundation

Infrastructure

Three-layer infrastructure stack: data ingestion, quantitative computation, and on-chain execution. Each layer optimized for latency and reliability.

Data Layer

Market feeds, on-chain indexers, oracle networks

DexScreener

CoinGecko

BaseScan

Etherscan

Pyth Oracle

Chainlink

System Health API

Event Feed

Compute Layer

Quantitative analysis, ML models, scheduling

Python

NumPy

Pandas

Web3.py

FastAPI

APScheduler

Model Monitor

Alert Engine

Execution Layer

Smart contract routing, DEX protocols, infrastructure

Base Chain

Aerodrome

Uniswap V3

Alchemy RPC

ERC-8004

Docker

x402 Premium

Discord Linked Roles

Voice Events

Payments Scaffold

DATA → COMPUTE → EXECUTION  ·  latency target: < 200ms

24h success: --

Recent rows: --

refreshing feed...

Access Points

Connect to the System

Four access layers. One operating system for intelligence.

ONLINE

Telegram Bot

@MCINTOSHIbot

Live on Telegram. Market intel, alerts, on-chain analytics, and premium trading education.

→

GitHub

@McIntoshiS

Open source components, smart contract deployments, and agentic infrastructure on Base.

→

X / Twitter

@SMcintoshi

Signal bursts. Market commentary. Edge identification. No noise, no filler — execution-grade content only.

→

Discord

discord.gg/5jHsNcTcuP

Community command center for builders, traders, and high-signal updates from the MCINTOSHI ecosystem.

→

"Edge identified. Execution pending."