CFSE.ai

A methodology for security epistemology

Turn complex systems into testable security knowledge.

CFSE connects system structure, security invariants, falsifiable hypotheses, and execution evidence—so every conclusion can be inspected, challenged, and reused.

Explore the method See it applied
From systemto evidence
  1. 01
    World modelModel the system
  2. 02
    InvariantState what must hold
  3. 03
    ScenarioForm a falsifiable question
  4. 04
    ExplorationRun the comparison
  5. 05
    TracePreserve the evidence
  6. 06
    FindingChange what is known

The artifact chain is not paperwork. It is how a security claim keeps its provenance.

Model before conclusionEvidence before assertionKnowledge after the run

Two ways in

Start with the method. Or start with the evidence.

01Understand the method

Build a security model that can be challenged.

Learn how CFSE moves from system structure to falsifiable hypotheses, controlled exploration, and evidence-backed conclusions.

Follow the methodology
02See the methodology at work

Watch how CFSE structures a real-world exploit.

A CFSE retrospective applies the full artifact chain — world model, invariants, scenarios, traces, findings — to a publicly reported Unitree Go2 vulnerability. Original discovery by boschko.ca. Methodology demonstration by us.

Open the CFSE retrospective

The method

How security knowledge compounds

Each step changes what the next step can legitimately claim. Skip the model, and the hypothesis floats. Skip the trace, and the conclusion loses its evidence.

01

Model the system

Name the actors, trust boundaries, interfaces, state, and legitimate flows before deciding where the vulnerability must be.

What this producesA queryable world modelConcepts · Entry Points · Interactions · Flows
02

State what must hold

Turn security expectations into explicit invariants whose state can change as evidence arrives.

What this producesA precise security contractInvariants · Predicates
03

Form a falsifiable scenario

Ask one bounded question that targets a specific invariant and declares what enforcement or violation would look like.

What this producesA testable hypothesisScenarios · Generators
04

Execute and preserve evidence

Compare a legitimate baseline with the manipulated path, then retain the ordered observations that support the verdict.

What this producesEvidence with provenanceExplorations · Traces
05

Make the knowledge reusable

Convert what the experiment changed into a bounded finding, a repaired invariant, and regression work that survives this target.

What this producesSecurity knowledge that compoundsFindings · Patches · Updated invariants

CFSE Retrospective / Unitree Go2

See the methodology applied to a known vulnerability.

We took a publicly reported exploit and modeled it through CFSE — Concepts, Entry Points, Invariants, Scenarios, Traces, Findings. Not our discovery. Our methodology, demonstrated.

See the retrospective
Phone
Map
Write
Press
Root
Public evidenceConfirmed

The published report describes the tamper-to-root sequence.

CFSE structureModeled

We reconstructed how CFSE would have represented the sequence.

What CFSE addsVisible

CFSE makes authority handoffs and evidence boundaries inspectable.

Local reproductionNot executed

The current corpus has no live Go2 target or faithful harness.

Precise underneath

The artifact system

Every object has one job. Together they preserve the path from system understanding to remediation without confusing hypotheses for evidence.

Spec 1.0.0
  1. 01
    C · EP · I · F

    World Model

    What exists, and how does authority move?

  2. 02
    INV

    Invariant

    What must remain true?

  3. 03
    S

    Scenario

    What specific failure can we falsify?

  4. 04
    E

    Exploration

    How do baseline and attack paths differ?

  5. 05
    T

    Trace

    What happened, in what order?

  6. 06
    FD

    Finding

    What conclusion does the evidence support?

  7. 07
    PATCH

    Patch

    How is the failure repaired and prevented elsewhere?

Read the specification →Learn through the guides →

Put it to work

Choose the depth you need.

Start with the guides

Learn the method in progressive, practitioner-facing order.

Use the specification

Inspect normative artifact schemas, semantics, and traceability rules.

Train with CFSE

Build the practice through Attify programs and team delivery.

Apply it to a system

Bring CFSE into a concrete product, research target, or assurance workflow.