
How to Brief a Coding Agent Before It Writes a Line
Every session your agent starts from zero. Here's what actually needs to travel with the prompt, and why a bigger model won't fix it.
Posts from the Thor team on the context layer AI agents and teams can trust.

Every session your agent starts from zero. Here's what actually needs to travel with the prompt, and why a bigger model won't fix it.

An agent that reads a stale ticket does exactly what it says, wrong parts included. See what sourced, reversible, governed context actually stops.