Mission Control
in the foundryThe control plane for multi-agent sessions.
Coordinates agent sessions, task packets, worktrees, prompts, and reviews across a run.
from agent-mission-control-lab · not yet public
Hekton Transformers
Gears that earn a public name.
Every Transformer starts as a lab inside the factory: a gear doing one job well. Once it proves itself, it graduates into a standalone public repo, a Transformer that works standalone as well as integrated into Hekton. Nothing here is vapourware; a lab either is a Transformer yet, or it isn't.
Interactive · The assembly
Every Transformer is a port on the Hekton OS core. Promoted ones are plugged in; the rest are still in the foundry, waiting to dock. Select one to inspect it.
●— plugged in○- - in the foundry
Select a Transformer above to see its dock status and where it came from.
In the Foundry
Not public yet
These labs are still proving themselves inside the factory. They'll move up here as they earn a Transformer name.
Agentic Infra
in the foundryCloud infra from a manifest, gated by a human apply.
Provisions cloud infrastructure from manifests with guardrails: Terraform plans behind a human apply gate.
from agentic-infra-lab · not yet public
Architecture Map
in the foundryA living map of the system, so nothing goes uninspected.
Maps the shape of the system and how its components fit, so the whole factory stays inspectable.
from architecture-map-lab · not yet public
Blog Factory
in the foundryTurns project work into publishable technical narratives.
A reusable engine that turns project work into reviewed, interactive, publishable technical narratives.
from blog-factory-lab · not yet public
Brand Signal
in the foundryKeeps public signal honest about the work underneath.
Keeps the public brand and social signal aligned with the actual work being done.
from brand-signal-lab · not yet public
Engine Gateway
in the foundryRoutes each job to the model that should handle it.
Model routing policy and the gateway in front of the LLMs: which model handles which job.
from engine-gateway-lab · not yet public
Local LLM
in the foundryA local-first inference baseline, cost-aware by default.
A local model baseline: Ollama, a model registry, and cost-aware local-first inference.
from local-llm-lab · not yet public
Obsidian Brain
in the foundryStructured memory for agents, built on Obsidian.
Obsidian as an intelligent second brain: structured memory, local embeddings, durable project context.
from obsidian-brain-lab · not yet public
Tekton Chronicle
in the foundryWhere build sessions become essays and field notes.
The writing room behind this journal: where build sessions become essays, artefacts, and field notes.
from tekton-chronicle · not yet public