Human reviewers
Use the site to understand what a risk check covers, what it does not cover, and which public artifact to read first.
Human view
Financial Agent Reliability Layer
TRADING4AI is not presenting itself as a signal shop or backtest platform in this phase. The public site explains how Action Preflight, Evidence Cards, and Reliability Corpus help humans and agents review risky financial actions and claims before trust gets overstated.
What exists now
Pages-only public delivery, shaped for people and machines.
The current public layer is static HTML plus static JSON artifacts. It is meant to be readable by humans, crawlable by search engines, and retrievable by AI systems without pretending to be a live judgment backend.
Action PreflightEvidence CardsReliability CorpusPages-only public delivery
Boundary
Static first. Verifiable first. Runtime later.
This phase focuses on public understanding, discovery, and sample delivery. No live backtesting, no Tencent backend runtime, and no request-time user judgment are implied by the public pages.
Who this is for
One public layer, three kinds of readers.
AI agents
Use the same pages and downloadable JSON to retrieve bounded service contracts, sample evidence shapes, and review limits.
Builders and operators
Use the public layer to align internal workflows around source refs, timestamps, limitation notes, and safer publication language.
Core services
Three public services explain the current product line.
These are the core public surfaces that search engines, AI systems, and humans should understand first.
Action Preflight
Pre-action review for payments, wallet transfers, subscriptions, and other high-risk financial moves.
Evidence Cards
Source-bound packaging with retrieval time, limitations, safe restatements, and blocked claims.
Reliability Corpus
Static warning patterns, assumption-drift notes, and claim-safety collections for repeated review work.
Recommended start
Start with the strongest public pack for the risk shape.
Most first-time human readers should begin with a concrete scenario pack before browsing the broader catalog.
Counterparty Screening Pack
Best first stop when a payment, transfer, or beneficiary change is close to action.
Claim Safety Pack
Best first stop when an agent is about to publish, promote, or restate a financial claim.
Crypto Signal Subscription Pack
Best next stop when the action involves guaranteed-return language, crypto payment, or brokerage access before trust is earned.
Wallet Transfer to Unverified Counterparty Pack
Best next stop when a crypto wallet destination is weakly verified and the transfer would be difficult to reverse once sent.
Public Materials Index
Use this when the task is still broad and the full static scenario catalog is the better starting point.
Public boundaries
What the public site does and does not do.
Pages-only public delivery
Every public surface in this phase is static HTML plus static JSON artifacts, built for retrieval and explanation.
No live backend judgment
The public pages do not accept user-submitted verification jobs, do not score live actions on the server, and do not imply autonomous execution.
No backtesting product in this phase
The site may discuss failure patterns and verification boundaries, but it is not advertising live user backtests or a strategy-performance leaderboard.
Machine discovery
Public machine links remain first-class.
- Reliability Home: /agent-verification?lang=en
- Current Capabilities: /current-capabilities?lang=en
- Action Preflight: /agent-verification/action-preflight?lang=en
- Evidence Cards: /agent-verification/evidence-cards?lang=en
- Reliability Corpus: /agent-verification/reliability-corpus?lang=en
- Counterparty screening before payment: /agent-verification/materials/counterparty-screening-pack?lang=en
- Claim Safety Pack: /agent-verification/materials/claim-safety-pack?lang=en
- Crypto signal subscription before payment: /agent-verification/materials/crypto-signal-subscription-pack?lang=en
- Wallet Transfer to Unverified Counterparty Pack: /agent-verification/materials/wallet-transfer-unverified-counterparty-pack?lang=en
- Machine Bridge: /llms?lang=en
- For Agents: /for-agents?lang=en
- Discovery Registry: /reliability/public-discovery-registry.json
- llms.txt: /llms.txt
- Agent card: /.well-known/agent.json
Suggested path
How a human should read the public layer
1. Start at Reliability Home to understand the three public services.
2. Open the service page that matches the task shape.
3. Open the strongest matching scenario pack before trusting any summary or artifact.
4. Use /llms only when you want to inspect the machine-facing discovery layer itself.
Human page FAQ
Common questions from first-time human readers
Question
What should a human read first on TRADING4AI?
Start with Reliability Home to understand the three public services, then open the strongest matching scenario pack, such as Counterparty Screening Pack, Claim Safety Pack, Crypto Signal Subscription Pack, or Wallet Transfer to Unverified Counterparty Pack, before treating any artifact as trustworthy.
Question
Should I treat these pages as live financial approval or compliance clearance?
No. These pages are public explanation and retrieval surfaces only. They do not provide live financial approval, user-specific compliance clearance, guaranteed safety, or autonomous execution.
Start here
Start with the public reliability layer before asking for deeper delivery.
Read the public services and priority packs first. They are the current truth surface for what TRADING4AI can explain, publish, and hand off safely in phase one.