TrueNorth Agentic Brokerage Review: Future of AI-Assisted Trading?
TrueNorth MCP, Trading Agents and Crypto Research Explained
The definitive TrueNorth review for 2027. Analyse its AI market intelligence, crypto analytics, MCP, expert playbooks, trade setups, prediction markets and non-custodial trading agents.
Edition: 2027 Early Edition
Last reviewed: 3 August 2026
Summary
TrueNorth is building an AI-native trading workspace that connects research, live data, structured trade setups, contextual memory, programmatic interfaces and optional non-custodial execution.
Its product architecture can be divided into five layers.
|
Layer |
Function |
|
Market-data layer |
Aggregates crypto, derivatives, onchain, equities, commodities and prediction-market data |
|
Intelligence layer |
Uses AI and specialised tools to analyse data |
|
Playbook layer |
Runs multi-step institutional-style research workflows |
|
Interface layer |
Desktop, Telegram, MCP, API and CLI access |
|
Execution layer |
Optional agents route orders to third-party decentralized venues |
TrueNorth states that it integrates more than 30 data sources and produces structured, trader-oriented outputs rather than generic prose. Its MCP layer provides specialist tools and expert playbooks, while the optional agent layer can submit transactions under user-granted permissions.
Category verdicts
|
Category |
DN assessment |
|
Crypto-native AI research |
Excellent |
|
Derivatives intelligence |
Strong |
|
Prediction-market research |
Strong |
|
Structured trade setups |
Strong, but verification required |
|
MCP integration |
Excellent strategic potential |
|
Portfolio context |
Developing |
|
Autonomous execution |
Powerful but high risk |
|
Custody architecture |
Non-custodial |
|
Regulatory status |
Intelligence provider, not regulated broker |
|
Best use |
AI-assisted research and supervised agent workflows |
Users can join TrueNorth through Decentralised News with referral code 5KHZNK.
The Central Proposition
The first generation of AI trading products asked users to trust a signal.
The next generation asks users to delegate a workflow.
That workflow can include:
- Scanning the market
- Identifying an asset
- Researching the narrative
- Analysing price structure
- Measuring derivatives positioning
- Producing an entry, stop and target
- Monitoring the setup
- Routing the order
- Managing or closing the position
- Recording the result
TrueNorth’s ambition is to become the layer connecting these tasks.
The opportunity is substantial.
So is the risk.
A mistaken chatbot answer is inconvenient.
A mistaken autonomous transaction is irreversible.
The DN Agentic Trading Intelligence Score
|
Category |
Weight |
DN score |
Assessment |
|
Market-data breadth |
15% |
9.1/10 |
More than 30 sources across several markets |
|
Crypto-native depth |
15% |
9.2/10 |
Onchain, derivatives, token and market intelligence |
|
Structured analysis |
15% |
9.0/10 |
Visual outputs and trade-setup formatting |
|
Source and freshness discipline |
10% |
8.2/10 |
Strong architecture, but user verification remains necessary |
|
MCP and programmatic access |
10% |
9.3/10 |
MCP, API and CLI strategy |
|
Workflow memory |
10% |
8.8/10 |
Persistent context differentiates it from blank-slate chat |
|
Execution controls |
10% |
7.8/10 |
Opt-in, non-custodial and emergency stop |
|
Regulatory clarity |
5% |
6.4/10 |
Not a licensed investment or brokerage intermediary |
|
Subscription transparency |
5% |
6.8/10 |
Credit model disclosed, live pricing remains changeable |
|
Legal and counterparty protection |
5% |
6.0/10 |
Broad disclaimers and limited platform liability |
Overall DN score: 8.5/10
The score evaluates product structure and workflow value.
It does not measure trading accuracy or profitability.
TrueNorth Product Map
|
Product component |
Purpose |
Primary user |
|
Desktop application |
Conversational research and visual analysis |
Active traders |
|
Market scanners |
Opportunity discovery |
Traders seeking new setups |
|
Specialist tools |
Fast access to individual metrics |
Analysts |
|
Expert playbooks |
Multi-source research reports |
Investors and traders |
|
MCP |
Brings TrueNorth into supported AI clients |
Advanced AI users |
|
API |
Programmatic data and platform integration |
Developers |
|
CLI |
Terminal-based interaction and agent management |
Technical users |
|
Agent execution |
Routes automated orders to third-party venues |
Advanced traders |
|
Credits |
Measures compute-intensive usage |
Free and paid users |
|
Referral programme |
Promotional user acquisition |
Existing and new users |
What TrueNorth Is Not
TrueNorth is not:
- A conventional brokerage account
- A centralized cryptocurrency exchange
- A futures commission merchant
- A registered investment adviser
- A fiduciary
- A guarantee of best execution
- A custodian
- A market maker
- An insurer against agent mistakes
The current legal terms identify TrueNorth as a market-intelligence and analytics service with optional non-custodial execution. Advent AI states that it does not control the exchanges, smart contracts, liquidity pools or order books to which agents can route orders.
Corporate and Legal Structure
|
Item |
Current public information |
|
Brand |
TrueNorth |
|
Contracting entity in current terms |
Advent AI Limited |
|
Incorporation stated in terms |
Cayman Islands |
|
Other public documentation reference |
ADVENT AI PTE. LTD. |
|
Investment-adviser registration |
None stated |
|
Broker-dealer registration |
None stated |
|
CFTC registration |
None stated |
|
Custody |
Non-custodial agent model |
|
Governing relationship |
Platform terms and third-party venue terms |
The presence of differing public corporate references does not prove a legal problem, but it is a reason to confirm the latest contracting entity before paying for a plan or granting execution authority.
The TrueNorth Data Layer
TrueNorth documentation identifies more than 30 real-time or near-real-time sources.
Named examples include:
- CoinGecko
- DeFiLlama
- Hyperliquid
- Polymarket
- CoinAnk
- Centralized exchanges
- Additional market and analytics providers
The value is not simply the number of integrations.
It is the ability to place different forms of evidence into one workflow.
|
Data category |
Example use |
|
Spot prices |
Confirm current market level |
|
Technical indicators |
Analyse momentum and structure |
|
Order books |
Estimate spread and liquidity |
|
Funding |
Measure long or short crowding |
|
Open interest |
Identify position expansion |
|
Liquidations |
Locate possible squeeze zones |
|
Onchain flows |
Observe exchange and protocol activity |
|
TVL and fees |
Measure protocol adoption |
|
Token unlocks |
Identify future supply |
|
Prediction markets |
Observe event probabilities |
|
News |
Identify catalysts |
|
Equity data |
Compare crypto and traditional markets |
|
Commodity data |
Analyse macro relationships |
Specialist Tools vs Expert Playbooks
|
Feature |
Specialist tool |
Expert playbook |
|
Purpose |
Retrieve one metric |
Run a complete analysis |
|
Speed |
Faster |
Longer workflow |
|
Inputs |
Focused question |
Broader objective |
|
Data sources |
Usually one or a few |
Several in parallel |
|
Output |
Specific answer |
Structured report |
|
Best use |
Known research need |
Uncertain or complex decision |
|
Example |
“Show ETH funding” |
“Should I buy ETH?” |
TrueNorth’s playbook layer is strategically important.
A generic chatbot requires the user to know which data to request.
A playbook can decide which tools should be called and in what sequence.
The danger is that orchestration can hide the quality of individual inputs.
A comprehensive report is only as reliable as:
- Its timestamps
- Data sources
- Labels
- Assumptions
- Liquidity context
- Counterarguments
Current Expert Playbook Categories
|
Playbook |
Main output |
|
Comprehensive Analysis |
Technical, derivative and narrative research |
|
Technical Trading Setup |
Entry, stop, target and risk-to-reward |
|
Tokenized Stock Research |
Fundamental and market analysis for tokenized equities |
|
Derivatives Analysis |
Funding, open interest and liquidation structure |
|
Prediction-Market Analysis |
Mechanics, probabilities and catalysts |
TrueNorth’s derivative templates include liquidation heatmaps, squeeze assessments, smart-money positioning and stop-hunt analysis.
The TrueNorth Trade-Setup Model
A TrueNorth output may contain:
|
Field |
Purpose |
|
Direction |
Long or short thesis |
|
Entry |
Proposed execution level |
|
Stop loss |
Thesis invalidation or risk limit |
|
Take profit |
Intended exit |
|
Reward-to-risk |
Price-distance comparison |
|
Position-sizing guidance |
Suggested exposure framework |
|
Chart |
Visual explanation |
|
Supporting indicators |
Evidence for the setup |
The output is precise in form.
It remains uncertain in outcome.
The DN Trade Setup Verification Test
|
Verification |
Required question |
|
Freshness |
When was the setup generated? |
|
Price source |
Which venue or index supplied the entry? |
|
Timeframe |
Is the setup intraday, swing or position trading? |
|
Spread |
What is the current bid and ask? |
|
Depth |
Can the intended size execute? |
|
Funding |
What is the carrying cost? |
|
Open interest |
Is the trade crowded? |
|
News |
Is an event approaching? |
|
Stop |
Can it execute before liquidation? |
|
Correlation |
Does the portfolio already hold the same risk? |
|
Fees |
What is the complete round-trip cost? |
|
Counter-thesis |
What evidence makes the setup wrong? |
A trade should not be delegated to an agent until every material question has an answer.
Mathematical Reward-to-Risk vs Real Probability
Assume:
- Entry: $100
- Stop: $97
- Target: $109
The displayed reward-to-risk ratio is:
$9 potential reward ÷ $3 potential risk = 3:1
This does not mean the trade has a 75% probability of success.
It does not account for:
- Win rate
- Gap risk
- Slippage
- Funding
- Fees
- Liquidation
- Partial fills
- Correlated portfolio losses
- Sudden news
TrueNorth’s own terms state that the ratio is a mathematical expression of distance rather than a probability forecast.
TrueNorth MCP
MCP allows an AI application to call external tools through a standardized protocol.
TrueNorth uses MCP to expose its market data, analytical tools and playbooks to supported AI clients.
|
MCP capability |
Potential use |
|
Price data |
Retrieve current market information |
|
Technical metrics |
Request RSI, MACD or support |
|
Derivatives data |
Funding, OI and liquidations |
|
Onchain metrics |
TVL, flows and activity |
|
Token economics |
Unlock and vesting analysis |
|
Prediction markets |
Retrieve market probabilities |
|
Expert playbooks |
Run complete research workflows |
|
Persistent AI workflow |
Continue analysis inside an AI client |
The documentation provides setup paths for generating an MCP link and connecting to Claude or ChatGPT.
MCP Security
|
Risk |
Required defence |
|
Leaked MCP token |
Store privately and revoke promptly |
|
Public repository exposure |
Never embed credentials in code |
|
Excessive permissions |
Use the minimum scope |
|
Prompt injection |
Restrict tool actions |
|
Unverified data request |
Require source and timestamp |
|
Automated bulk extraction |
Follow platform quotas |
|
Hidden downstream use |
Review connected AI applications |
|
Subscription overuse |
Monitor credit consumption |
The terms make users responsible for API keys, MCP tokens and CLI sessions. Programmatic access can be throttled or terminated if limits are exceeded or credentials are misused.
API and CLI Architecture
|
Interface |
Main role |
|
REST API |
Structured requests and account functions |
|
WebSocket API |
Streaming market or account updates |
|
MCP |
AI-client tool access |
|
CLI |
Terminal interaction and agent management |
TrueNorth can change, deprecate or discontinue programmatic interfaces.
Users building critical systems should therefore plan for:
- Version changes
- Tool-definition changes
- Rate-limit changes
- Authentication changes
- Outages
- Credit-price changes
- Fallback data sources
Non-Custodial Agent Architecture
The agent layer separates custody from automation.
|
Component |
Responsible party |
|
Wallet keys |
User |
|
Wallet software |
Third-party provider |
|
Agent configuration |
User |
|
TrueNorth analysis |
Advent AI systems |
|
Order routing |
TrueNorth agent |
|
Market execution |
Third-party venue |
|
Smart contract |
Third-party venue |
|
Settlement |
Blockchain or venue |
|
Resulting P&L |
User |
TrueNorth says it never holds the private keys or seed phrase.
The user can nevertheless lose funds because previously granted wallet permissions can allow an agent to transact within authorised parameters.
Agent Permissions Ladder
|
Permission level |
Capability |
DN classification |
|
0 |
Research only |
Lower operational risk |
|
1 |
Alerts and monitoring |
Conservative |
|
2 |
Draft orders requiring confirmation |
Controlled |
|
3 |
Small automatic orders |
Elevated |
|
4 |
Continuous autonomous execution |
High |
|
5 |
Broad wallet and venue authority |
Extreme |
Most users should begin at Level 0 or Level 1.
Level 3 should use a separate, limited-capital wallet.
Levels 4 and 5 require professional-grade controls.
Agent Risk Matrix
|
Risk |
How it can appear |
|
Data error |
Agent acts on stale or incorrect inputs |
|
Hallucination |
AI invents a catalyst or relationship |
|
Latency |
Market moves before execution |
|
Slippage |
Order fills far from expected price |
|
Compounding error |
Several bad trades occur before intervention |
|
Margin shortfall |
Agent lacks collateral to manage positions |
|
Liquidation |
Leveraged position is forcibly closed |
|
Smart-contract exploit |
Execution venue loses funds |
|
Oracle failure |
Position uses inaccurate price data |
|
Venue downtime |
Agent cannot close or modify |
|
Wallet compromise |
Permissions are abused |
|
Gas spike |
Transaction costs rise |
|
Regulatory change |
Venue access becomes restricted |
|
Emergency-stop delay |
Orders continue during deactivation |
TrueNorth’s terms specifically identify smart-contract, execution, continuous-operation, compounding-error, venue, gas and regulatory risks.
The DN Agent Guardrail Standard
|
Guardrail |
Minimum policy |
|
Asset allowlist |
Only explicitly approved markets |
|
Position notional |
Hard maximum per asset |
|
Leverage |
Hard maximum |
|
Per-trade loss |
Defined percentage of equity |
|
Daily loss |
Agent stops after threshold |
|
Order types |
Explicitly permitted |
|
Market orders |
Human confirmation above threshold |
|
New asset |
Human approval |
|
New venue |
Human approval |
|
Wallet approval |
Minimum amount and contract scope |
|
Data age |
Maximum permissible timestamp age |
|
Liquidity |
Minimum depth and maximum spread |
|
Stop loss |
Required before or with entry |
|
Monitoring |
Real-time alerts |
|
Emergency stop |
Tested |
|
Independent revoke |
Wallet or venue control outside TrueNorth |
Emergency Stop Reality
TrueNorth provides an emergency-stop function.
The terms warn that:
- Deactivation may not be instantaneous.
- Network and processing conditions affect timing.
- Transactions may be submitted during the interval.
- Already confirmed transactions cannot be reversed.
The safest agent design assumes the emergency stop can fail or arrive late.
The primary defence is limited authority.
Subscription and Credit Economy
|
Subscription element |
Current terms |
|
Free tier |
Available |
|
Paid tiers |
Available |
|
Billing |
Monthly recurring |
|
Credits |
Included by plan |
|
Top-ups |
May be purchased |
|
Unused monthly credits |
Normally expire |
|
Transferability |
Credits cannot be transferred |
|
Cash value |
None |
|
Refunds |
Generally unavailable |
|
MCP/API consumption |
Can differ from platform usage |
|
Pricing changes |
Permitted with notice for existing subscribers |
|
Referral credits |
Campaign dependent |
Measuring TrueNorth’s True Cost
Subscription price alone is incomplete.
Use:
TrueNorth true monthly cost = subscription + top-up credits + connected AI subscription + trading fees + builder fees + gas + funding + slippage + losses caused by automation errors
A low subscription cost can still support an expensive operating workflow.
Credit-Efficiency Framework
|
User behaviour |
Credit efficiency |
|
Focused specialist query |
High |
|
Well-defined playbook |
High |
|
Repeated vague prompts |
Low |
|
Duplicate analysis |
Low |
|
Running agents without need |
Low |
|
Using MCP for planned workflows |
Potentially high |
|
Continuous unnecessary scanning |
Low |
|
Saving reusable prompt templates |
High |
Referral Programme
The user-provided Decentralised News referral route is:
Join TrueNorth with code 5KHZNK
The general referral terms do not promise a fixed reward.
Promotional credits may:
- Change
- Be capped
- Expire
- Be withdrawn
- Have no cash value
Duplicate accounts, self-referrals and automated referral generation can lead to forfeiture and account restriction.
TrueNorth vs General-Purpose AI
|
Feature |
TrueNorth |
General-purpose chatbot |
|
Crypto-specific data |
Integrated |
Tool dependent |
|
Derivatives metrics |
Built into specialist tools |
Requires external sources |
|
Onchain analytics |
Integrated |
Requires connectors |
|
Prediction markets |
Integrated |
Requires browsing or API |
|
Structured setups |
Core workflow |
Prompt dependent |
|
Charts |
Native emphasis |
Varies |
|
Persistent trading context |
Platform feature |
Product dependent |
|
MCP |
Native product layer |
Can connect to third parties |
|
Execution agents |
Optional |
Requires external infrastructure |
|
Main risk |
Automation and data assumptions |
Hallucination and stale information |
TrueNorth vs Traditional Charting Platform
|
Feature |
TrueNorth |
Traditional charting |
|
Conversational analysis |
Native |
Limited or separate |
|
Charting |
Included in outputs |
Core strength |
|
Onchain data |
Integrated |
Often limited |
|
Derivatives intelligence |
Integrated |
Depends on provider |
|
AI playbooks |
Core product |
Usually add-on |
|
Trade setup generation |
Automated |
User or script driven |
|
Order execution |
Agent-routed where enabled |
Broker integration dependent |
|
Independent judgement |
Still required |
Still required |
TrueNorth vs No-Code Bot
|
Feature |
TrueNorth agent |
Traditional no-code bot |
|
Inputs |
AI analysis plus configured rules |
Fixed conditions |
|
Adaptability |
Higher |
Lower |
|
Determinism |
Potentially lower |
Higher |
|
Explainability |
Depends on output |
Rule logic visible |
|
Hallucination risk |
Present |
Limited |
|
Compounding-error risk |
High |
Present |
|
Best use |
Supervised research-led automation |
Repeatable tested rules |
Trader-Type Fit
|
User |
Best TrueNorth use |
|
Beginner |
Research and education only |
|
Long-term investor |
Fundamental, narrative and portfolio research |
|
Swing trader |
Technical and derivative playbooks |
|
Perpetual trader |
Funding, OI and liquidation analysis |
|
Prediction trader |
Market mechanics and catalyst review |
|
Onchain analyst |
Wallet, protocol and smart-money research |
|
Quant developer |
API, MCP and CLI |
|
Professional trader |
Controlled agents and multi-source monitoring |
|
Content creator |
Research acceleration and visual analysis |
Daily TrueNorth Research Desk
|
Stage |
Action |
|
Market state |
Scan trends, funding, OI and liquidations |
|
Catalyst state |
Review news, unlocks and economic events |
|
Narrative state |
Identify accelerating and weakening themes |
|
Candidate state |
Shortlist a maximum of three ideas |
|
Research state |
Run comprehensive playbook |
|
Verification state |
Inspect original evidence |
|
Execution state |
Size position outside AI output |
|
Approval state |
Human accepts or rejects |
|
Monitoring state |
Alerts and agent limits |
|
Review state |
Record outcome and process errors |
A Complete TrueNorth Prompt Framework
Context
State:
- Asset
- Venue
- Timeframe
- Portfolio objective
- Existing exposure
- Maximum risk
Evidence
Request:
- Current timestamp
- Primary sources
- Data providers
- Price
- Funding
- OI
- Liquidations
- Onchain activity
- Liquidity
Thesis
Require:
- Bull case
- Bear case
- Neutral case
- Unknowns
Setup
Request:
- Entry
- Stop
- Targets
- Reward-to-risk
- Fees
- Slippage
- Funding
Failure
Require:
- Invalidation condition
- Venue failure scenario
- Liquidity failure scenario
- News-event risk
Authority
State:
- Whether execution is manual
- Agent notional cap
- Allowed assets
- Permitted order types
- Human-confirmation threshold
Copy-Ready Advanced Prompt
“Act as a market-intelligence analyst, not an autonomous decision-maker.
Analyse [ASSET] on [VENUE] for a [TIMEFRAME] trade.
State the exact timestamp and source for all time-sensitive data.
Use technical structure, volume, spread, order-book depth, funding, open interest, liquidation distribution, relevant onchain activity and verified catalysts.
Separate confirmed facts from inference.
Provide the strongest bull case, bear case and neutral interpretation.
Produce a hypothetical entry, stop and target only after documenting the assumptions.
Estimate fees, slippage and funding.
Define the evidence that invalidates the setup.
Do not assume wallet identity, venue eligibility, legal status or execution quality.
Do not submit any transaction.
End with the human checks required before approval.”
Decentralised News Proprietary Tool
DN TrueNorth Trade Setup & Agent Guardrail Validator
Validate an AI-generated entry, stop and target, calculate risk-sized position limits, and score the permissions and safeguards surrounding an autonomous trading agent.
1. Validate the trade setup
Use the actual venue fee, expected slippage and leverage. The tool sizes the position from acceptable account loss rather than from available buying power.
2. Score the agent guardrails
Tick only controls that are configured, tested and enforceable. A verbal intention is not a guardrail.
DN TrueNorth Trade Setup and Agent Guardrail Validator
The tool contains two linked modules.
Trade Setup Validator
It calculates:
- Account risk
- Maximum planned loss
- Stop distance
- Target distance
- Risk-sized notional
- Required margin
- Quantity
- Modelled fees
- Slippage
- Carrying cost
- Net target profit
- Net reward-to-risk
- Agent notional-cap usage
Agent Guardrail Score
It evaluates:
- Asset allowlist
- Position cap
- Leverage cap
- Per-trade loss
- Daily loss
- Order permissions
- Human approval
- Wallet scope
- Venue eligibility
- Data freshness
- Emergency stop
- Monitoring
- Independent revocation
Score interpretation
|
Score |
DN status |
|
Below 60 |
Agent disabled |
|
60 to 79 |
Supervised use only |
|
80 to 94 |
Small controlled pilot |
|
95 to 100 |
Strong controls, not guaranteed safety |
Due-Diligence Questions
Before enabling execution, ask:
- Which execution venues are currently supported?
- Which smart contracts receive wallet approval?
- Can approvals be capped by amount?
- Can assets be allowlisted?
- Can leverage be capped?
- Can market orders be disabled?
- Can a daily-loss stop be enforced?
- Is human approval available above a threshold?
- How quickly does emergency deactivation work?
- Can permissions be revoked directly through the wallet?
- What happens during a TrueNorth outage?
- What happens during a venue outage?
- How are failed transactions handled?
- How are agent decisions logged?
- Can the complete activity history be exported?
- Which data source triggered each order?
- What credits are consumed during continuous operation?
- Which corporate entity provides the service?
- Which law governs disputes?
- Is the intended activity lawful in the user’s jurisdiction?
Frequently Asked Questions
Is TrueNorth an AI trading bot?
It is broader than a conventional bot. It combines research, data, playbooks, market monitoring and optional agent execution.
Is TrueNorth a regulated brokerage?
No. Its terms state that it is not a registered broker-dealer, investment adviser, commodity trading adviser or futures commission merchant.
Does TrueNorth custody cryptocurrency?
The agent-execution architecture is described as non-custodial.
Where do TrueNorth trades execute?
Agents route orders to independent third-party protocols, DEXs, liquidity pools or order books.
Can TrueNorth reverse an agent trade?
No. Confirmed onchain or venue transactions are generally irreversible.
Can the agent continue after I press stop?
There can be a delay before deactivation becomes effective.
Can an AI-generated stop guarantee my loss?
No. Fast markets, gaps, liquidity problems and execution delays can create a worse outcome.
Does TrueNorth guarantee accurate data?
No. Its terms disclaim guarantees concerning accuracy, completeness and timeliness.
Can I use TrueNorth without trading?
Yes. The intelligence and analytics features are intended to function independently of agent execution.
Does TrueNorth integrate with ChatGPT?
The MCP documentation provides a connection workflow for supported ChatGPT use.
What is a TrueNorth playbook?
A playbook is a multi-step research workflow that calls several tools and synthesises the results.
Does TrueNorth analyse Polymarket?
Yes. Prediction-market intelligence is one of its documented categories.
Does TrueNorth analyse crypto options?
Yes. It covers options flow, volatility, open interest, Greeks and related market structure.
How does TrueNorth charge?
It uses free and paid subscription tiers with usage credits. Exact pricing and consumption should be checked live.
Do TrueNorth credits expire?
Monthly credits normally expire at the end of the billing cycle unless a plan specifically states otherwise.
What is the referral code?
Use code 5KHZNK through the TrueNorth Decentralised News referral link.
Final 2027 Verdict
TrueNorth represents a credible vision of where trading software is moving.
The conventional trader assembles:
- A charting platform
- An onchain terminal
- A derivatives dashboard
- A news service
- A prediction-market interface
- A portfolio tracker
- An automation tool
- A general-purpose AI model
TrueNorth attempts to place these functions inside one contextual intelligence layer.
That can reduce research friction.
It can also concentrate operational dependence.
The platform’s strongest qualities are:
- Crypto-native intelligence
- Multi-asset breadth
- Specialist data tools
- Expert playbooks
- Structured visual outputs
- MCP strategy
- Contextual continuity
- Non-custodial execution architecture
Its most important limitations are:
- AI uncertainty
- Third-party venue dependence
- Wallet-permission risk
- Agent compounding-error risk
- Limited regulatory role
- Broad liability disclaimers
- Credit and subscription complexity
- The possibility that emergency controls act too late
The correct TrueNorth adoption ladder is:
- Use it as a research assistant.
- Verify every material output.
- Create reusable playbooks.
- Connect MCP with read-only intent.
- Build alerts.
- Validate setups independently.
- Test a simulated or tiny manual trade.
- Create enforceable agent guardrails.
- Use a separate limited-capital wallet.
- Run a closely monitored pilot.
- Expand only after reviewing a meaningful sample.
- Retain human authority over exceptional actions.
The real question is not whether AI can place a trade.
It is whether the system can explain, constrain, monitor and survive the trades it places.
TrueNorth is one of the more ambitious attempts to answer that question.
Users can join TrueNorth through Decentralised News with referral code 5KHZNK.
Affiliate Disclosure
The TrueNorth link included in this publication is a referral link. Decentralised News may receive compensation when an eligible reader registers, subscribes or completes qualifying activity.
Affiliate relationships do not guarantee inclusion, a favourable assessment or profitable results.
Risk Disclaimer
This publication is for educational and informational purposes only. It does not constitute financial, investment, legal, tax or trading advice.
AI-generated content can be inaccurate, incomplete, stale or misleading. Automated agents can submit unintended orders, experience latency, interact with vulnerable smart contracts and create losses faster than a person can intervene.
Users are responsible for reviewing every permission, risk parameter and execution venue. Never provide a seed phrase, never grant broader wallet authority than necessary and never expose capital that cannot be lost.
