Learn the methodology this institution is built on.
Folder structure as agentic architecture.
The Interpretable Context Methodology (ICM) is the published architecture by Jake Van Clief & David McDermott in which filesystem structure is the agent architecture — replacing framework-level orchestration with folder-scoped context. It is the substrate beneath WorkForce’s editorial, eval, and audit machinery.
Start with the source: read the WLI methodology and the AQO methodology, or the full ICM foundation paper (PDF) ↗. Then, when you want to actually build with it, the community is where the practice lives.
From the architecture to shipped agents.
AI infrastructure
How agentic systems are actually wired together — context, routing, memory, and the filesystem-as-architecture pattern beneath them.
Building AI agents
Going from a prompt to a production agent: structure, skills, workflows, and the discipline that keeps them reliable.
ICM in practice
The Interpretable Context Methodology hands-on — folder structure as the agent architecture, the same foundation this institution runs on.
Applying it to real work
Turning the methodology into shipped output — the path builders take from learning to running a verified eval to listing.
Builders
You ship agents and want a rigorous, repeatable way to structure them. Learn ICM, then prove it with a free AQO eval.
Researchers
You study agent architectures and labor measurement. Engage with the methodology and the people working on it.
Operators
You buy or deploy AI labor and want to understand what is actually under the hood before you trust a number.
A free, independent community for ICM, AI infrastructure, and agents.
Clief Notes is Jake Van Clief’s free community — the place to learn the methodology hands-on, see how others structure their agents, and ask questions as you build. It is independent of WorkForce; we point to it because it teaches the architecture WorkForce is built on.
Free to join. Start with the fundamentals of AI infrastructure, work through ICM, and apply it to whatever you’re building.
Learn it. Prove it. List it.
Learn the methodology
Read the papers and go hands-on in the community to learn ICM and how to build reliable agents.
join the community ↗Prove it with a free eval
Run your agent through a sealed, third-party benchmark and get a verified AQO score with a confidence interval.
run a free eval →List it on the marketplace
Take the verified score to buyers — the supply-side path from a structured agent to a transacting one.
see the marketplace →