overarching LLM standards and strategies for the org
This field has a context window of its own: what you knew a month ago isn't wrong; it has simply scrolled out of attention. The only way to stay in the window is to keep showing up.
This is the org's shared home for you and your agents, a guiding light for anyone growing their understanding, and an open forum for how we build with AI in a landscape that reinvents itself every few months. Staying current here isn't a nice-to-have. It's a discipline, and one we practice together.
Maximize value per win, where a win costs tokens plus human attention. Get this right and cost takes care of itself.
- Cutting tokens caps capability along with spend
- Maximizing value per win improves cost and quality
- Every tool choice, workflow, and practice we discuss gets measured against TTV
- Documentation: Ratified principles and patterns; draft stack guidance
- Governance: The official process for proposing, refining, and ratifying content
- Discussions: The living forum where ideas are shaped before they solidify
- Disagreement is welcome. Nobody sees the whole board alone
If you've found a tool that changed how you work, a workflow that stopped paying for itself, or a claim that deserves scrutiny, bring it here.
This entire initiative requires constant tuning and contribution, stewardship will be a communal effort.