Skip to main content

Documentation Index

Fetch the complete documentation index at: https://kindling.birklid.com/llms.txt

Use this file to discover all available pages before exploring further.

ArticleApplications Layer

​
What Happens When You Let AI Trade Crypto for a Week?

A first-person experiment handing over a live crypto portfolio to an AI agent — an honest account of the results and what they reveal about AI trading in practice.

​
What it covers

This narrative case study documents a one-week experiment running an AI agent as an autonomous crypto portfolio manager. The author gives the agent real funds, real market access, and minimal guardrails — then reports what actually happened. The piece covers the agent’s decision-making patterns, where it succeeded (and why), where it failed (and why), and what the experience reveals about the current state of AI in autonomous trading contexts. Rather than benchmarks and paper trading results, this is a practitioner’s first-hand account — the kind of specific, messy, real-world data that’s harder to find than theoretical writeups. Worth reading alongside the more technical resources in this section for a grounded perspective on AI trading’s practical limitations.
Read this when you want to ground your understanding of AI trading agents in reality rather than benchmark performance — this is what actually happens when you remove the guardrails.

​
Read the article

Read: What Happens When You Let AI Trade Crypto for a Week? ↗

Full narrative case study.

How to trade with AI the right way

NexusTrade’s methodology guide — a counterpoint on doing it properly.

TradingAgents

Multi-agent framework for structured, explainable AI trading strategies.