Choose a stack¶
Translate requirements into experimental questions.
Choose the system boundary first¶
Do you need building blocks for an application you will design, or a complete coding environment around the model?
| Decision | Start with |
|---|---|
| I need orchestration, tools, state, approvals, or delegation APIs inside my application | Frameworks and SDKs |
| I need a coding agent that already owns prompts, tools, execution, permissions, persistence, and interaction | Coding agents and harnesses |
Do not combine them in one ranking. A harness owns many variables that the framework experiment deliberately holds fixed. The comparison gateway explains the boundary.
Start here¶
Write down provider/protocol, tool access, approvals, durability, latency and budget requirements. For frameworks, read the decision guide and feature matrix. For complete environments, use the coding-harness matrix. Check the profile evidence label, inspect the methodology, and reproduce the relevant arena before choosing.
Prerequisites: a small task you can describe, its permitted effects, and basic Python for executable examples. Reading the guides needs no API key.
Evidence and coverage¶
Existing offline comparisons establish mechanics. Native provider answer-quality rankings remain unavailable until repeated live scorecards are published.
Follow the concept and build path for runnable lessons and the ecosystem map for wider coverage. Return to all journeys.