False Systems® Talk with us ↗
False Systems / Computing as craftBerlin — Helsinki / 00

Delegate more.
Move faster.
Keep control.

One Operation connects local work, remote execution, and a customer cluster while keeping their boundaries distinct.
Fig. 01 / One work identity. Different places.

Humans.
Agents.
Shared work.

One control plane for humans, AI agents, and infrastructure.

Coordinate work across teams and environments. Keep identity, authority, and evidence connected from development to production.

Infrastructure for platform and security teams working with autonomous agents.

Explore the platform ↗Talk with us ↗
Control Plane coordinates / Habitat executes / Evidence establishesMeet the products ↓
01 / The problem

The work crosses boundaries.
Control doesn’t follow.

A developer asks an agent to fix a service. Code goes to a repository, tests run in CI, and a deployment uses cloud credentials to change a cluster.

Each tool understands its part. The relationship between the request, the permission to act, and the result can be lost between them.

  1. 01 / RequestHumanA task
  2. 02 / WorkAgentA session
  3. 03 / ChangeRepositoryA commit
  4. 04 / TestCIA job
  5. 05 / AccessCloudA credential
  6. 06 / DeployKubernetesA workload
One piece of work. Different identities, permissions, and records.

People reconstruct context by hand. Agents can duplicate work, permissions can exceed the task, and handoffs can lose decisions. When something fails, the supporting records are scattered.

02 / The approach

Keep people and agents
working together.

Everyone involved in a task needs to know what was requested, who is responsible, what has been approved, and what still needs to happen.

False Systems is designed to connect assignments, decisions, permissions, and results across tools and handoffs. Teams should be able to continue the task without reconstructing its history from separate conversations.

One shared taskRequest · Owner · Decisions · Results
PeopleAgentsTools & environments
Conceptual view: people and agents share an account of the task, even when they work through different tools.
03 / Products

Three responsibilities.
One system.

Products overview ↗

One piece of work · shared context across all three

01 / Coordinate

Control Plane

Many agents. One organization.

Coordinate work across agents, humans, and teams. Keep ownership, delegation, approvals, and handoffs in a common organizational context.

Explore Control Plane ↗
02 / Execute

Habitat

The work outlives the worker.

Run work across execution environments. Give the task an identity, bounded authority, and retained results independent of a particular worker.

Explore Habitat ↗
03 / Establish

Evidence

Evidence, not explanations.

Connect observed actions to their actors, grants, and history. Make the supporting records checkable and the uncertain outcomes visible.

Explore Evidence ↗

Product responsibilities, not a synchronous execution stack. Coordination and evidence publication are separate from the local execution path.

04 / Your existing tools

Keep your tools.
Connect the work.

Your agents can use different harnesses. Your organization still needs one way to coordinate their work.

An agent harness manages how an agent works. An organizational harness manages how agents and humans work together: shared context, ownership, dependencies, approvals, and handoffs.

Control Plane is designed to provide that layer across tools such as Claude Code, Codex, and Pi, and workflows involving GitHub, cloud infrastructure, and Kubernetes. These illustrate the intended scope, not a verified integration list.

Explore Control Plane ↗
Organization

Humans · Teams · Policies

Control Plane / Organizational harness

Shared work · Ownership · Delegation · Approvals · Handoffs

Coding agentCodex harness
Testing agentClaude Code harness
Investigating agentPi harness
HumanReview & approval

HabitatOperations · Workers · Hosts

EvidenceActors · Grants · Observations · Outcomes

Conceptual relationships, not verified integrations with the named harnesses. People and agents share organizational context; local execution does not pass through a central service for every action.
05 / In practice

An agent-driven
production change.

One request. Several workers. A human decision before the privileged step.

Illustrative workflow / Not a live connected demonstration

  1. A human requests a service change.

    Control Plane relates the request to its owner, intended result, and scope.

  2. A coding agent prepares the change.

    The agent uses its own harness. Habitat associates execution with an Operation and its permitted resources.

  3. Other workers test and investigate.

    One runs tests; another examines an infrastructure dependency. Their work and handoffs stay related to the original request.

  4. A human reviews the privileged step.

    Production access requires a separate decision. Approval concerns the specific action and scope, not unrestricted credentials for the agent.

  5. An authorized worker performs the permitted action.

    Habitat checks runtime authority. Evidence connects observations to the worker, grant, and work; an unconfirmed external outcome stays unconfirmed.

Control Plane coordinates decisions. Habitat manages execution. Evidence supports checking the account. This example describes the intended product relationship, not universal integrations or production-wide enforcement.

06 / A different responsibility

Running agents is only
part of the problem.

False Systems addresses the relationship between organizational work, authority, and outcomes across existing systems. The categories overlap; each still has a job to do.

Agent harnesses
Manage individual agent execution, context, and tools.
Orchestrators
Manage workload scheduling and desired state.
Observability tools
Collect and query system signals.
Security tools
Enforce particular policies and boundaries.
07 / Who it is for

Built for teams giving agents
real responsibility.

Platform engineering
Coordinate work across development environments, CI, infrastructure, and clusters without tying the task to one agent session.
Security engineering
Make delegated authority explicit and examine the evidence associated with sensitive actions, including gaps and uncertain outcomes.
Engineering leadership
Understand shared work, ownership, dependencies, progress, and handoffs across people and agents.
08 / Research & Journal

Thinking in public.
Checking the assumptions.

The Journal ↗
Platform thinking / Habitat

The Process Is Not the Work ↗

What should survive when the worker disappears? An investigation of work identity, authority, and responsibility, informed by systems history.

Independent open-source experiment / Sykli

Pipelines don’t rot ↗

A separate experiment in execution infrastructure on the BEAM. Read the question and its limits in the Journal, then inspect the public source ↗.

Journal essays explain the thinking; independent research tests particular ideas. Neither establishes complete commercial-platform guarantees. Syvä and Sykli are distinct projects, not required platform components. Read the historical execution research ↗

09 / A direct conversation

Bring a
real workflow.

Working with agents across development, infrastructure, or production? Let’s examine where coordination, authority, and evidence become difficult.

Talk with False Systems ↗

Independent systems thinking. Berlin & Helsinki. ↗