threat-modeling.sgit.ai / graph

Threat models as vault-native graph data

Threat models are graph data, not prose — that is the whole argument, and this page has to avoid contradicting it by being a pile of markdown. Each published model gets a JSON twin: nodes (Asset, Threat, Vulnerability, Mitigation, Actor, Incident), edges (targets, mitigates, is-instance-of, crosses-trust-boundary), and a per-finding status.

The platform's pitch to an AppSec team is not "we have better threat modelling than X"; it is "we have a clean home for all your security artifacts, with tools that work directly on top of them."
— the AppSec mini-tools strategy brief, 16 May 2026

Eight frictions, and the vault answer

An AppSec engineer juggles eight to fifteen tools, each with its own data home. The tool choice is usually fine. The data home is the friction — and each of these maps directly onto a threat model's own lifecycle.

The questionThe vault answer
Where does the threat model live?In a vault — versioned, owned, exportable
How do I share it with auditors?Read-only vault link, time-limited
How do I bring in new data?Vault content appends; tools read from the vault
How do I run offline / air-gapped?Vault collections are portable; tools run on local SG/Compute
How do I correlate findings across tools?All tools write the same vault structure; findings cross-reference
How do I version-control security artifacts?The vault commit history is the version control
How do I prove what was reviewed when?Vault commits are cryptographically signed; auditors verify
How do I share without losing control?Vault sharing with revocation

Several of these are the mandatory-disclosure paper's requirements arriving as product features: a signed, dated, revocable, auditor-verifiable artefact is precisely the "reliable signal" that paper says the market lacks.

The artefact schema

A threat-modelling vault holds: system descriptions, threat lists (STRIDE-categorised), attack trees, DREAD scores, mitigations, Gherkin test cases, and evidence. Two of those carry this page's real point:

the strongest answer to "static document"

Gherkin test cases

A mitigation expressed as an executable scenario is a threat model CI can check — not a claim re-read at the next audit, but a check that runs on every change.

the artefact that closes the loop

Evidence

What turns a claim into a disclosure: the thing that lets a reader verify the model rather than trust it.

The tool pattern, and the step this site holds itself to

The productisation template: find a strong open-source tool, host it on shared infrastructure, add a UI and vault integration, contribute back. The first named target is StrideGPT — actively developed, LLM-integrated, supporting STRIDE plus the OWASP LLM Top 10 and MAESTRO-inspired pattern detection, with two intended modes: pure client-side for air-gapped use, and ephemeral-compute-backed for teams.

Stated intention, not a partnership. No contact with StrideGPT's maintainer has been confirmed as of this release. This site does not imply a relationship that does not exist — see comms question Q5 on /admin/comms.html.

Eight further mini-tools sit on the same substrate — SBOM analysis, dependency scanning, secrets detection, OWASP Top 10 assessment, compliance evidence, policy management, pentest artefacts, vulnerability disclosure management. Only the threat-modelling one belongs on this site; the rest are a pointer. "Contribute back" is a commitment this site makes visible rather than re-argues — open-source.sgit.ai holds the founder's position on open source itself.

What is real today versus what is argued

The gap between the 2025 papers' vision and today's implementation is this site's main honesty risk, so it is drawn as an explicit table rather than left implicit.

Real, nowArgued, not built
The ThreatModCon vault: 11 linked models, 51 nodes, 179 threats, live and offline-capableThe continuous pipeline that keeps models current as code changes
Vault primitives: versioning, signing, revocable read-only sharing, sub-vaults, append lanesAutomated CVE-to-component linking that instantly flags a new threat
MGraph-DB as a memory-first graph storeThe multi-ontology overlay (STRIDE × ATT&CK × OWASP) running as a live query over a production model
Seven written threat models, one validated against codeGherkin-executable mitigations checked in CI
StrideGPT identified as the first integration targetThe integration itself

An agent reading this site should be able to tell, for any capability named on it, whether it can use it today or is being told a direction. That is not a courtesy — it is the same falsifiability standard /validated/ applies to a single finding, applied here to the platform's own claims.

Where this site stops and a sibling starts

TopicOwnerThis site's part
Wardley maps as a techniquewardley-maps.sgit.aiOnly the five threat-prioritisation walkthroughs, as an application
G³, MGraph-DB, graph theorygraphs.sgit.aiOnly threat-model-shaped graphs; link out for the substrate
ISO 27001, EU AI Act, GDPRstandards.sgit.aiOnly framework-overlay-as-graph-edge; standards text stays there
Risk registers, acceptancerisks.sgit.aiOnly the risk_decision field as it appears in the security schema
Vault mechanics, keyspki.sgit.ai / sgit.aiOnly what a threat model needs to say about them

The rule that keeps this clean: this site owns the act of modelling threats. If material is about the thing being modelled rather than the modelling, it belongs to a sibling. The full network map →