I built an autonomous treasury agent, then let a code review bot find every way it could lose money
I just submitted TreasuryForge to the WeMakeDevs × TrueFoundry Agent Harness Hackathon, an autonomous agent that manages a simulated treasury across cash, crypto, and NSE equities, built entirely on TrueForge , TrueFoundry's agent harness. This isn't a writeup about the idea. It's about what actually broke, what a code review bot caught before it shipped, and what I learned wiring a real approval…
This story details the process of building an autonomous treasury management system called TreasuryForge, which was submitted to a hackathon. The system manages a simulated treasury across cash, crypto, and NSE equities using TrueForge, a TrueFoundry agent harness. The hackathon judges required TrueForge to interact with tools, run code in a sandbox, and stop for human approval before executing any trades.
The system includes a pre-trade analysis, sandboxed risk checks, an approval gate, automated self-review, and a second agent that audits the first agent's history. The author discusses various issues found during code reviews, including an approval gate bypass, double execution of trades, race conditions, incorrect test cases, and limitations of free-tier LLMs.
These problems highlight the importance of thorough testing and careful implementation when creating autonomous systems.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
