Navigable state
Loop state lives as addressable content, not a flat transcript. Agents resume from the exact section, symbol, row, run, or decision that matters.
Harness-neutral loop substrate
The substrate agentic loops run against. Assemble content and capabilities into pods, then let any compatible harness resume, verify, and compose work against the same navigable state.
Collections
Nodes
one pod surface, scoped by roleRoles
draft + testinspect + citedeploy + run checksEngine vs substrate
Loop state lives as addressable content, not a flat transcript. Agents resume from the exact section, symbol, row, run, or decision that matters.
A pod is the workspace abstraction: collections provide data sources, nodes provide capabilities, and the loop runs against that assembled surface.
Any compatible harness can target the same substrate. The agent is swappable; the pod state and operation surface compound.
Pods and nodes
Each MCP connection uses one pod. Collections hold the content the loop can address; nodes provide capabilities the loop can invoke for that content.
native:skillsIndexes qualified SKILL.md files across the pod so agents can discover, read, pin, and curate reusable playbooks.
native:ciAdds test() so draft changes can be checked before they become committed state.
native:runnerAdds run() for builds, smoke tests, and commands on a machine you connect over SSH.
Navigable resumable state
Lynko turns natural document and code structure into handles any agent harness can unfold, cite, edit, verify, and compose. The MCP connector is already scoped to its pod, so calls operate on that pod's collections and nodes.
lynko-org[loop-substrate-thesis.md].section("Moat")
lynko-platform[src/server.go].expand("Handler").draft.edit(...)
skills.search("dashboard review")
lynko-platform.test(targets="api")
lynko-platform.run("npm run smoke", machine="staging")Early access