Published research, experiments, and engineering experience help us decide what to build and test next. Explore the questions guiding WTK and the evidence that could change our approach.
WTK brings research into a governed lifecycle for agents and agent teams.
Define the goal, assemble the package, and set its limits.
Test the exact execution form and retain failures and limitations.
A human authorizes where the qualified version may run.
Compare a candidate with the accepted version before promotion.
Outside research, development records, and WTK findings, with each publication’s evidence and limitations kept distinct.
An agent can fail at a task and report that failure accurately. A different problem arises when it receives clear failure evidence but tells the next person or system that the work is complete.
Some instructions tell every agent how to work together. Others tell one agent how to do its job. Revising them separately can make the intended change and its consequences easier to inspect.
An agent may notice a problem before anyone asks for help. It still needs to decide whether to wait, notify someone, prepare a response or carry out an authorized action.
WTK is a governed AI agent factory for building, qualifying, deploying, and improving portable agents and agent teams. Its evidence-driven improvement loops can propose and test changes, while promotion remains independently governed and human-authorized.
A governed, self-improving AI agent factory that progressively automates the building, qualification, deployment, operation, and improvement of portable agents and agent teams while keeping authority, evidence, and promotion independently governed.
Outside research and practical observations help us identify questions and design tests. We use WTK to investigate those questions, then document the results, limitations, and any changes to the approach.
Outside findings are not evidence that WTK has passed the same test. Research may lead to a design change, a proposed experiment, or no change when the evidence is insufficient.
Relevant research, practical results, and alternative approaches can inform this work. Contributions are reviewed privately by a human. Public use requires separate permission; submission does not result in automatic publication.
