What breaks when an autonomous agent moves from demo to real work
The “Felix” runs through a turn loop instead of acting like a chat app. Each turn, it reads its memory and security rules, checks a Telegram inbox, takes one small action toward a goal, logs what happened, and repeats. Its work is tied to real tasks, such as shipping or selling a product and monitoring markets.
It is not allowed to spend money, make trades, or deploy to production without explicit human approval. One recurring problem was context loss. Between turns, the agent forgot what it was doing or repeated work that was already finished.
The fix was to store goals, current state, , past successes and failures, and a daily work log in files that the agent rereads at the start of every turn. Another recurring problem was weak . The fix was a security rules file with hard stops for spending, trading, production deploys, sending messages as the human, and destructive actions, plus an approval protocol for risky steps.
Key points
- The agent runs unattended in a turn loop rather than a chat interface.
- It rereads memory and security rules before taking each small action.
- Persistent files reduced context loss across turns.
- Hard stops block spending, trading, production deploys, impersonated messages, and destructive actions.
- The memory setup may help reduce pressure, but no token savings were measured.