Products

Overview

This is still not a product in the usual sense.

It is a named container for a pattern that is now showing up repeatedly in real work: multiple agent roles are useful only when handoffs, verification duties, and publication boundaries are explicit enough to survive execution.

Status

Incubation stage.

This page now rests on real supporting material rather than pure speculation, but it is still too early to present as a finished toolkit. The question is no longer whether the pattern is interesting. The question is which parts become stable enough to package.

Current Evidence

The pattern now has several public anchors:

• the methodology page for how FAE chooses and publishes experiments
• the autonomous website maintenance loop experiment
• the research analyst + operator split experiment
• the agent handoff reliability trial
• the Rick/Greg coordination model that shaped the split operationally

That is enough to justify incubation language, but not enough to imply a mature product.

What Might Eventually Belong Here

• role definitions and boundaries
• analyst-to-operator handoff formats
• verification checklists
• publication safety classes
• shared-memory or context-transfer patterns
• lightweight tooling that reduces coordination overhead

Availability

No standalone public toolkit yet.

For now, this page should be read as an incubation record: a recurring internal pattern with enough experimental support to name, but not enough stability to sell or package aggressively.

Overview

This is a candidate internal tool, not a launched product.

The underlying problem is real: once experiments span multiple models, providers, and workflows, it becomes harder to understand total operating cost from memory or scattered logs. A cost meter would exist to turn that drift into something inspectable.

Status

Concept stage. No public artifact is linked yet.

The value of this page, for now, is to record a possible incubation path rather than imply that a finished system already exists.

Why It Might Exist

• compare actual cost across agent workflows
• make model-routing tradeoffs more explicit
• show where “cheap” experiments stop being cheap in aggregate

What Would Need To Exist Before This Becomes Real

• a repeatable internal use case
• a concrete data model for cost capture
• at least one working artifact or spec worth linking publicly