PrismDevRelOpen the app
Case studies: Security and threat model

Overview

Case studies

UORust

How UORust was modeledArchitecture walkthroughDesign views in OPM and UMLFeatures and journeysSecurity and threat modelQuality model (ISO/IEC 25010)Behavior and proofsSystem and releaseExtension facetsCommand referenceSlides and video scriptRecord explorerllms.txt and Markdown twins

Prism-DevRel

How Prism-DevRel was modeledArchitecture walkthroughDesign views in OPM and UMLFeatures and journeysSecurity and threat modelQuality model (ISO/IEC 25010)Behavior and proofsExtension facetsSlides and video scriptRecord explorerllms.txt and Markdown twins

Rule 30

How Rule 30 was modeledArchitecture walkthroughSecurity and threat modelQuality model (ISO/IEC 25010)Behavior and proofsSlides and video scriptRecord explorerllms.txt and Markdown twins

Verdant Resonance

How Verdant Resonance was modeledBehavior and proofsExtension facetsSlides and video scriptRecord explorerllms.txt and Markdown twins

Foundry

How Foundry was modeledBehavior and proofsSlides and video scriptRecord explorerllms.txt and Markdown twins

// case study · UORust

Security and threat model

The assets of UORust, its threats, risks and controls, and the risks it accepts on purpose.

Subject: UORust at commit afd5cea950a8, 331 facet records, source digest f3f743b6834adaff.

Security and threat model records by kindAcceptedRisk: 1 records; ApplicationSecurityControl: 13 records; Asset: 6 records; ControlSet: 1 records; ModeledAsset: 6 records; ModeledControl: 13 records; ModeledThreat: 11 records; Risk: 11 records; SecurityActivity: 13 records; Threat: 11 records; ThreatMapping: 10 records; VerificationMeasurement: 12 recordsAcceptedRisk1ApplicationSecurityControl13Asset6ControlSet1ModeledAsset6ModeledControl13ModeledThreat11Risk11SecurityActivity13Threat11ThreatMapping10VerificationMeasurement12
Security and threat model records by kind. The 108 records this page teaches, one bar for each kind. Each kind has its own section below.
The data in this figure
records
AcceptedRisk1
ApplicationSecurityControl13
Asset6
ControlSet1
ModeledAsset6
ModeledControl13
ModeledThreat11
Risk11
SecurityActivity13
Threat11
ThreatMapping10
VerificationMeasurement12

The security facet holds ISO/IEC 27034 and ISO/IEC 27005 records: assets, threats, risks and controls. Its Production.ThreatModel rows give them names, STRIDE categories and evidence. Once any row is present, the SDK checks three things. Every threat has a control or an accepted risk (PP2019). Every control has evidence (PP2020). Every threat has a category (PP2021). See PrismPM SPEC section 2.3.

Security and threat model facet explains the facet for every Prism model, with the counts of every case study.

Every record

Every record 108 rows
IndexRecordKindName
215Security::acceptedPagesHostSubstitutionAcceptedRiskacceptedPagesHostSubstitution
216Security::activityArtifactClosureSecurityActivityactivityArtifactClosure
217Security::activityBoundedIngestSecurityActivityactivityBoundedIngest
218Security::activityCargoDenySecurityActivityactivityCargoDeny
219Security::activityConfinedOutputSecurityActivityactivityConfinedOutput
220Security::activityDeterministicOutputSecurityActivityactivityDeterministicOutput
221Security::activityForbidUnsafeSecurityActivityactivityForbidUnsafe
222Security::activityLockedBuildSecurityActivityactivityLockedBuild
223Security::activityNoBuildExecutionSecurityActivityactivityNoBuildExecution
224Security::activityOfflineSecurityActivityactivityOffline
225Security::activityPinnedWorkflowsSecurityActivityactivityPinnedWorkflows
226Security::activityRouterConformanceSecurityActivityactivityRouterConformance
227Security::activityTypedRefusalSecurityActivityactivityTypedRefusal
228Security::activityVerifiedRouteTableSecurityActivityactivityVerifiedRouteTable
229Security::assetCiAssetassetCi
230Security::assetEmittedModelAssetassetEmittedModel
231Security::assetPagesSiteAssetassetPagesSite
232Security::assetReleaseBinaryAssetassetReleaseBinary
233Security::assetSubjectSourceAssetassetSubjectSource
234Security::assetSupplyChainAssetassetSupplyChain
235Security::controlArtifactClosureApplicationSecurityControlcontrolArtifactClosure
236Security::controlBoundedIngestApplicationSecurityControlcontrolBoundedIngest
237Security::controlCargoDenyApplicationSecurityControlcontrolCargoDeny
238Security::controlConfinedOutputApplicationSecurityControlcontrolConfinedOutput
239Security::controlDeterministicOutputApplicationSecurityControlcontrolDeterministicOutput
240Security::controlForbidUnsafeApplicationSecurityControlcontrolForbidUnsafe
241Security::controlLockedBuildApplicationSecurityControlcontrolLockedBuild
242Security::controlNoBuildExecutionApplicationSecurityControlcontrolNoBuildExecution
243Security::controlOfflineApplicationSecurityControlcontrolOffline
244Security::controlPinnedWorkflowsApplicationSecurityControlcontrolPinnedWorkflows
245Security::controlRouterConformanceApplicationSecurityControlcontrolRouterConformance
246Security::controlTypedRefusalApplicationSecurityControlcontrolTypedRefusal
247Security::controlVerifiedRouteTableApplicationSecurityControlcontrolVerifiedRouteTable
248Security::mappingBuildScriptExecutionThreatMappingmappingBuildScriptExecution
249Security::mappingCraftedSourceExhaustionThreatMappingmappingCraftedSourceExhaustion
250Security::mappingDependencyCompromiseThreatMappingmappingDependencyCompromise
251Security::mappingDisputedProjectionThreatMappingmappingDisputedProjection
252Security::mappingMisroutedFactThreatMappingmappingMisroutedFact
253Security::mappingPathEscapeThreatMappingmappingPathEscape
254Security::mappingSourceExfiltrationThreatMappingmappingSourceExfiltration
255Security::mappingTamperedPagesThreatMappingmappingTamperedPages
256Security::mappingUnsafeMemoryThreatMappingmappingUnsafeMemory
257Security::mappingWorkflowTokenAbuseThreatMappingmappingWorkflowTokenAbuse
258Security::measurementArtifactClosureVerificationMeasurementmeasurementArtifactClosure
259Security::measurementBoundedIngestVerificationMeasurementmeasurementBoundedIngest
260Security::measurementCargoDenyVerificationMeasurementmeasurementCargoDeny
261Security::measurementDeterministicOutputVerificationMeasurementmeasurementDeterministicOutput
262Security::measurementForbidUnsafeVerificationMeasurementmeasurementForbidUnsafe
263Security::measurementLockedBuildVerificationMeasurementmeasurementLockedBuild
264Security::measurementNoBuildExecutionVerificationMeasurementmeasurementNoBuildExecution
265Security::measurementOfflineVerificationMeasurementmeasurementOffline
266Security::measurementPinnedWorkflowsVerificationMeasurementmeasurementPinnedWorkflows
267Security::measurementRouterConformanceVerificationMeasurementmeasurementRouterConformance
268Security::measurementTypedRefusalVerificationMeasurementmeasurementTypedRefusal
269Security::measurementVerifiedRouteTableVerificationMeasurementmeasurementVerifiedRouteTable
270Security::modeledAssetCiModeledAssetCI pipeline
271Security::modeledAssetEmittedModelModeledAssetEmitted Prism model
272Security::modeledAssetPagesSiteModeledAssetPages site
273Security::modeledAssetReleaseBinaryModeledAssetRelease binary
274Security::modeledAssetSubjectSourceModeledAssetSubject source code
275Security::modeledAssetSupplyChainModeledAssetSupply chain
276Security::modeledControlArtifactClosureModeledControlArtifact closure
277Security::modeledControlBoundedIngestModeledControlBounded ingest
278Security::modeledControlCargoDenyModeledControlDependency policy
279Security::modeledControlConfinedOutputModeledControlConfined output
280Security::modeledControlDeterministicOutputModeledControlDeterministic output
281Security::modeledControlForbidUnsafeModeledControlNo unsafe code
282Security::modeledControlLockedBuildModeledControlLocked build
283Security::modeledControlNoBuildExecutionModeledControlNo build of the subject but its MIR dump
284Security::modeledControlOfflineModeledControlNo network
285Security::modeledControlPinnedWorkflowsModeledControlPinned workflows
286Security::modeledControlRouterConformanceModeledControlRouter conformance
287Security::modeledControlTypedRefusalModeledControlTyped refusal
288Security::modeledControlVerifiedRouteTableModeledControlVerified route table
289Security::modeledThreatBuildScriptExecutionModeledThreatCode execution from the subject
290Security::modeledThreatCraftedSourceExhaustionModeledThreatCrafted source exhausting uorust
291Security::modeledThreatDependencyCompromiseModeledThreatCompromised dependency
292Security::modeledThreatDisputedProjectionModeledThreatDisputed projection
293Security::modeledThreatMisroutedFactModeledThreatMisrouted fact
294Security::modeledThreatPagesHostSubstitutionModeledThreatCompromised Pages host
295Security::modeledThreatPathEscapeModeledThreatWrite outside the output directory
296Security::modeledThreatSourceExfiltrationModeledThreatSource exfiltration
297Security::modeledThreatTamperedPagesModeledThreatTampered Pages artifact
298Security::modeledThreatUnsafeMemoryModeledThreatMemory corruption
299Security::modeledThreatWorkflowTokenAbuseModeledThreatWorkflow token abuse
300Security::riskBuildScriptExecutionRiskriskBuildScriptExecution
301Security::riskCraftedSourceExhaustionRiskriskCraftedSourceExhaustion
302Security::riskDependencyCompromiseRiskriskDependencyCompromise
303Security::riskDisputedProjectionRiskriskDisputedProjection
304Security::riskMisroutedFactRiskriskMisroutedFact
305Security::riskPagesHostSubstitutionRiskriskPagesHostSubstitution
306Security::riskPathEscapeRiskriskPathEscape
307Security::riskSourceExfiltrationRiskriskSourceExfiltration
308Security::riskTamperedPagesRiskriskTamperedPages
309Security::riskUnsafeMemoryRiskriskUnsafeMemory
310Security::riskWorkflowTokenAbuseRiskriskWorkflowTokenAbuse
311Security::threatBuildScriptExecutionThreatthreatBuildScriptExecution
312Security::threatCraftedSourceExhaustionThreatthreatCraftedSourceExhaustion
313Security::threatDependencyCompromiseThreatthreatDependencyCompromise
314Security::threatDisputedProjectionThreatthreatDisputedProjection
315Security::threatMisroutedFactThreatthreatMisroutedFact
316Security::threatModelControlSetControlSetUORust security controls
317Security::threatPagesHostSubstitutionThreatthreatPagesHostSubstitution
318Security::threatPathEscapeThreatthreatPathEscape
319Security::threatSourceExfiltrationThreatthreatSourceExfiltration
320Security::threatTamperedPagesThreatthreatTamperedPages
321Security::threatUnsafeMemoryThreatthreatUnsafeMemory
322Security::threatWorkflowTokenAbuseThreatthreatWorkflowTokenAbuse

AcceptedRisk (1)

ApplicationSecurityControl (13)

Asset (6)

ControlSet (1)

UORust security controls

Security::threatModelControlSet ControlSet #316 source

uorust reads a Cargo workspace and writes a Prism project. It runs locally, builds nothing it reads, opens no network connection and writes only below its output directory.

version
0.1.1
lastModified
2026-11-11T00:00:00Z

Taught by Security and threat model

ModeledAsset (6)

ModeledControl (13)

Artifact closure

Security::modeledControlArtifactClosure ModeledControl #276 source

prismpm verify replays the acceptance vectors in the built guest and checks the exact file closure of the browser artifact.

control
Security::controlArtifactClosure
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    prismpm --project app verify

Taught by Security and threat model

Bounded ingest

Security::modeledControlBoundedIngest ModeledControl #277 source

Every read is bounded in size and every walk in depth; a bound that is hit ends the run with a typed error.

control
Security::controlBoundedIngest
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    UE-06: Every way a run can end without a projection is a typed UorustError variant with a negative test

Taught by Security and threat model

Dependency policy

Security::modeledControlCargoDeny ModeledControl #278 source

cargo deny refuses advisories, unknown registries, wildcard requirements and licenses outside Apache-2.0, MIT and Unicode-3.0.

control
Security::controlCargoDeny
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    just vv deny

Taught by Security and threat model

Confined output

Security::modeledControlConfinedOutput ModeledControl #279 source

The emitter writes only relative paths it generated itself, each checked to stay below the output directory.

control
Security::controlConfinedOutput
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.review)
    reference
    crates/uorust/src/lib.rs:168 pub fn write(out: &Path, files: &BTreeMap<String, Vec<u8>>) -> Result<(), UorustError> {

Taught by Security and threat model

Deterministic output

Security::modeledControlDeterministicOutput ModeledControl #280 source

Output is a pure function of the workspace and its commit time, written in a fixed order, so a dispute is settled by re-running uorust.

control
Security::controlDeterministicOutput
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    UE-02: Two projections of the same workspace are byte-identical, for the fixture and for this workspace

Taught by Security and threat model

No unsafe code

Security::modeledControlForbidUnsafe ModeledControl #281 source

#![forbid(unsafe_code)] in every uorust crate makes any unsafe block a compile error.

control
Security::controlForbidUnsafe
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    crates/uorust/src/main.rs:3 #![forbid(unsafe_code)]
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    crates/uorust/src/lib.rs:12 #![forbid(unsafe_code)]

Taught by Security and threat model

Locked build

Security::modeledControlLockedBuild ModeledControl #282 source

Every build and test runs with --locked against the committed Cargo.lock.

control
Security::controlLockedBuild
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    just vv lint and test run cargo with --locked

Taught by Security and threat model

No build of the subject but its MIR dump

Security::modeledControlNoBuildExecution ModeledControl #283 source

uorust project, check and prove read the workspace through cargo metadata and syn only and never compile the subject; uorust mir alone compiles it with cargo, running its build scripts and macros as any build of it does, to dump the MIR the lift translates, and it is run only on a workspace the reviewer would build.

No build of the subject but its MIR dump 2 entries
control
Security::controlNoBuildExecution
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.review)
    reference
    crates/uorust/src/ingest.rs:578 let parsed = match syn::parse_file(&text) {
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.review)
    reference
    crates/uorust/src/mir/mod.rs:194 .env("RUSTC_BOOTSTRAP", "1")
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    UE-01: uorust projects the fixture workspace into exactly the committed Prism project, which lexlean fmt, lexlean lock and prismpm check accept

Taught by Security and threat model

No network

Security::modeledControlOffline ModeledControl #284 source

uorust links no network client, and every cargo command it runs (cargo metadata, and the MIR dump's cargo clean and cargo rustc) runs with --offline against the subject's lock file.

control
Security::controlOffline
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    cargo deny check over Cargo.lock
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.review)
    reference
    no line of crates/uorust/src names std::net outside a doc comment (checked by scripts/self-model/facets.py)

Taught by Security and threat model

Pinned workflows

Security::modeledControlPinnedWorkflows ModeledControl #285 source

Workflows pin every action by commit, check out with persist-credentials false and are themselves pinned by template.lock.

control
Security::controlPinnedWorkflows
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    xtask audit-bootstrap

Taught by Security and threat model

Router conformance

Security::modeledControlRouterConformance ModeledControl #286 source

The Rust router replays every vector of model/routes.toml, the same vectors the verified guest replays.

control
Security::controlRouterConformance
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    UE-05: The engine's router equals model/routes.toml route for route and answers every router vector of that table exactly

Taught by Security and threat model

Typed refusal

Security::modeledControlTypedRefusal ModeledControl #287 source

Admission never defaults: an ambiguous, unknown or incomplete fact becomes an open claim with its reason.

control
Security::controlTypedRefusal
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    UE-03: Admission refuses an ambiguous call, a feature without acceptance and a missing MSRV with typed refusals recorded as open claims, and emits no record for them

Taught by Security and threat model

Verified route table

Security::modeledControlVerifiedRouteTable ModeledControl #288 source

The route table is a LexLean function proved total and ambiguity-refusing by decide, and checked by the Lean kernel.

control
Security::controlVerifiedRouteTable
evidence
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.proof)
    reference
    Routes::routeTableIsTotal
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.proof)
    reference
    Routes::routeTableRefusesAmbiguity
  • Evidence
    kind
    (ctor Production.ThreatModel::EvidenceKind.gate)
    reference
    prismpm --project app verify

Taught by Security and threat model

ModeledThreat (11)

Code execution from the subject

Security::modeledThreatBuildScriptExecution ModeledThreat #289 source

A build script or procedural macro in the subject workspace runs on the reviewer's machine while it is modeled.

threat
Security::threatBuildScriptExecution
category
(ctor Production.ThreatModel::Stride.elevationOfPrivilege)

Taught by Security and threat model

Misrouted fact

Security::modeledThreatMisroutedFact ModeledThreat #293 source

A fact becomes the wrong record, or an ambiguous fact is resolved by a default, so the model misstates the source.

threat
Security::threatMisroutedFact
category
(ctor Production.ThreatModel::Stride.tampering)

Taught by Security and threat model

Write outside the output directory

Security::modeledThreatPathEscape ModeledThreat #295 source

A crafted crate name makes the emitter write outside the directory given by --out.

threat
Security::threatPathEscape
category
(ctor Production.ThreatModel::Stride.tampering)

Taught by Security and threat model

Tampered Pages artifact

Security::modeledThreatTamperedPages ModeledThreat #297 source

The published page or guest is altered so it routes facts differently from the verified table.

threat
Security::threatTamperedPages
category
(ctor Production.ThreatModel::Stride.tampering)

Taught by Security and threat model

Risk (11)

riskBuildScriptExecution

Security::riskBuildScriptExecution Risk #300 source

index
0
assetIndex
4
threatIndex
0
likelihood
(ctor Foundation.Sec::Likelihood.medium)
impact
(ctor Foundation.Sec::Impact.critical)

Taught by Security and threat model

riskCraftedSourceExhaustion

Security::riskCraftedSourceExhaustion Risk #301 source

index
1
assetIndex
1
threatIndex
1
likelihood
(ctor Foundation.Sec::Likelihood.medium)
impact
(ctor Foundation.Sec::Impact.minor)

Taught by Security and threat model

riskDependencyCompromise

Security::riskDependencyCompromise Risk #302 source

index
2
assetIndex
5
threatIndex
2
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.critical)

Taught by Security and threat model

riskDisputedProjection

Security::riskDisputedProjection Risk #303 source

index
3
assetIndex
1
threatIndex
3
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.minor)

Taught by Security and threat model

riskMisroutedFact

Security::riskMisroutedFact Risk #304 source

index
4
assetIndex
1
threatIndex
4
likelihood
(ctor Foundation.Sec::Likelihood.medium)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

riskPagesHostSubstitution

Security::riskPagesHostSubstitution Risk #305 source

index
5
assetIndex
2
threatIndex
5
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

riskPathEscape

Security::riskPathEscape Risk #306 source

index
6
assetIndex
3
threatIndex
6
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

riskSourceExfiltration

Security::riskSourceExfiltration Risk #307 source

index
7
assetIndex
4
threatIndex
7
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.critical)

Taught by Security and threat model

riskTamperedPages

Security::riskTamperedPages Risk #308 source

index
8
assetIndex
2
threatIndex
8
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

riskUnsafeMemory

Security::riskUnsafeMemory Risk #309 source

index
9
assetIndex
3
threatIndex
9
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

riskWorkflowTokenAbuse

Security::riskWorkflowTokenAbuse Risk #310 source

index
10
assetIndex
0
threatIndex
10
likelihood
(ctor Foundation.Sec::Likelihood.low)
impact
(ctor Foundation.Sec::Impact.major)

Taught by Security and threat model

SecurityActivity (13)

Threat (11)

ThreatMapping (10)

VerificationMeasurement (12)

measurementArtifactClosure

Security::measurementArtifactClosure VerificationMeasurement #258 source

index
0
controlIndex
0
passed
true

Taught by Security and threat model

measurementBoundedIngest

Security::measurementBoundedIngest VerificationMeasurement #259 source

index
1
controlIndex
1
passed
true

Taught by Security and threat model

measurementDeterministicOutput

Security::measurementDeterministicOutput VerificationMeasurement #261 source

index
3
controlIndex
4
passed
true

Taught by Security and threat model

measurementForbidUnsafe

Security::measurementForbidUnsafe VerificationMeasurement #262 source

index
4
controlIndex
5
passed
true

Taught by Security and threat model

measurementNoBuildExecution

Security::measurementNoBuildExecution VerificationMeasurement #264 source

index
6
controlIndex
7
passed
true

Taught by Security and threat model

measurementPinnedWorkflows

Security::measurementPinnedWorkflows VerificationMeasurement #266 source

index
8
controlIndex
9
passed
true

Taught by Security and threat model

measurementRouterConformance

Security::measurementRouterConformance VerificationMeasurement #267 source

index
9
controlIndex
10
passed
true

Taught by Security and threat model

measurementTypedRefusal

Security::measurementTypedRefusal VerificationMeasurement #268 source

index
10
controlIndex
11
passed
true

Taught by Security and threat model

measurementVerifiedRouteTable

Security::measurementVerifiedRouteTable VerificationMeasurement #269 source

index
11
controlIndex
12
passed
true

Taught by Security and threat model

explanationEvaluatecase studyMarkdown

Words on this page 4
Prism model
A description of an app, written in LexLean, that covers every facet of it and that the Lean kernel checks before the app is built.
facet
One view of a model, such as its architecture, its security or its quality. Each facet holds records.
record
One entry in a facet: a component, a threat, a measure or a journey step, with the place in the source it came from.
SDK
Software development kit. Here, the PrismPM container image that builds, checks and verifies Prism models.