consensus

ICorrelationTracker.computeCorrelationMatrix

Implementation of

getCorrelation()

getCorrelation(agentA, agentB): number | undefined;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:159

Get correlation between two specific agents. Returns undefined if insufficient data.

Parameters

agentA

string

agentB

string

Returns

number | undefined

Implementation of

ICorrelationTracker.getCorrelation

getStats()

getStats(): CorrelationTrackerStats;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:224

Get statistics about the correlation tracker.

Returns

CorrelationTrackerStats

Implementation of

ICorrelationTracker.getStats

hasSufficientData()

hasSufficientData(agentIds): boolean;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:196

Check if there is sufficient correlation data for a set of agents.

Parameters

agentIds

readonly string[]

Returns

boolean

Implementation of

ICorrelationTracker.hasSufficientData

identifyIndependentSubsets()

identifyIndependentSubsets(): readonly IndependentSubset[];

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:169

Identify groups of agents that vote independently.

Returns

readonly IndependentSubset[]

Implementation of

ICorrelationTracker.identifyIndependentSubsets

recordProposalVotes()

recordProposalVotes(
   proposalId, 
   votes, 
   outcome, 
   context?
): void;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:103

Record votes from multiple agents for the same proposal.

Parameters

proposalId

string

votes

ReadonlyMap<string, { conditions?: string[]; confidence: number; decision: "approve" | "reject" | "abstain"; findings?: { claim: string; gate: { named_assertion: string; reread_cited_line: "failed" | "skipped" | "passed"; ruled_out_language_non_issue: "failed" | "skipped" | "passed"; traced_call_path: "failed" | "skipped" | "passed"; }; location: string; severity: "critical" | "high" | "low" | "medium"; summary: string; }[]; reasoning: string; rejectionCategories?: ( | "YAGNI" | "DRY_VIOLATION" | "OVER_ENGINEERING" | "SCOPE_CREEP" | "SECURITY_RISK" | "MISALIGNED" | "INSUFFICIENT_EVIDENCE")[]; selectedOption?: string; timestamp?: string; }>

outcome

"rejected" | "approved"

context?

CorrelationRecordContext

Returns

void

Implementation of

ICorrelationTracker.recordProposalVotes

recordVote()

recordVote(
   agentId, 
   vote, 
   outcome, 
   context?
): void;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:79

Record a vote and its outcome for correlation tracking.

Parameters

agentId

string

vote

conditions?

string[] = ...

confidence

number = ...

decision

"approve" | "reject" | "abstain" = VoteDecisionSchema

findings?

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

string = ...

rejectionCategories?

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

string = ...

timestamp?

string = ...

outcome

"rejected" | "approved"

context?

CorrelationRecordContext

Returns

void

Implementation of

ICorrelationTracker.recordVote

setCurrentModelPins()

setCurrentModelPins(modelPins): void;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:74

Select the active model partition for each pinned role.

Parameters

modelPins

ReadonlyMap<string, string>

Returns

void

Implementation of

ICorrelationTracker.setCurrentModelPins

HigherOrderVotingStrategy

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:311

Higher-order voting strategy for integration with VotingStrategyFactory. Wraps OWVoting to provide IVotingStrategy interface.

Extends

OWVoting

Implements

IVotingStrategy

Constructors

Constructor

new HigherOrderVotingStrategy(options?): HigherOrderVotingStrategy;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:312

Parameters

options?

OWVotingOptions = {}

Returns

HigherOrderVotingStrategy

Overrides

OWVoting.constructor

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning";

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:72

Implementation of

IVotingStrategy.algorithm

Inherited from

OWVoting.algorithm

Methods

aggregate()

aggregate(votes, tracker?): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:199

Full pipeline: estimate correlation, compute result.

tracker is OPTIONAL (#3173): when omitted, the instance uses the tracker injected at construction (OWVotingOptions.tracker), letting higher-order voting be reused as a building block without threading the tracker through every call. A per-call tracker still wins. Throws if neither is available.

Parameters

votes

tracker?

Returns

Inherited from

OWVoting.aggregate

aggregateWithCorrelation()

aggregateWithCorrelation(votes, correlationMatrix): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:92

Aggregate votes using Bayesian correlation-aware method.

Parameters

votes

correlationMatrix

Returns

OWVoting.aggregateWithCorrelation

Inherited from

calculateOutcome()

calculateOutcome(votes, _weights?): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:87

IVotingStrategy implementation for integration with ConsensusEngine.

Parameters

votes

Map<string, { conditions?: string[]; confidence: number; decision: "approve" | "reject" | "abstain"; findings?: { claim: string; gate: { named_assertion: string; reread_cited_line: "failed" | "skipped" | "passed"; ruled_out_language_non_issue: "failed" | "skipped" | "passed"; traced_call_path: "failed" | "skipped" | "passed"; }; location: string; severity: "critical" | "high" | "low" | "medium"; summary: string; }[]; reasoning: string; rejectionCategories?: ( | "YAGNI" | "DRY_VIOLATION" | "OVER_ENGINEERING" | "SCOPE_CREEP" | "SECURITY_RISK" | "MISALIGNED" | "INSUFFICIENT_EVIDENCE")[]; selectedOption?: string; timestamp?: string; }>

_weights?

Map<string, number>

Returns

IVotingStrategy.calculateOutcome

Implementation of

Inherited from

OWVoting.calculateOutcome

computeISP()

computeISP(votes, independentSubsets): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:152

Compute result using Independent Subset Partition method.

Parameters

votes

independentSubsets

readonly IndependentSubset[]

Returns

Inherited from

OWVoting.computeISP

estimateCorrelation()

estimateCorrelation(tracker): CorrelationMatrix;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:148

Estimate correlation matrix from voting history.

Parameters

tracker

Returns

OWVoting.estimateCorrelation

Inherited from

getConfig()

getConfig(): HigherOrderVotingConfig;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:251

Get the current configuration.

Returns

HigherOrderVotingConfig

Inherited from

OWVoting.getConfig

NoAdapterError

Defined in: packages/nexus-agents/src/cli/voter-adapter-resolve.ts:25

Error thrown when no adapter is available and simulation is disabled.

Extends

Error

Constructors

Constructor

new NoAdapterError(message): NoAdapterError;

Defined in: packages/nexus-agents/src/cli/voter-adapter-resolve.ts:26

Parameters

message

string

Returns

NoAdapterError

Overrides

Error.constructor

Properties

cause?

optional cause?: unknown;

Defined in: node_modules/.pnpm/typescript@6.0.3/node_modules/typescript/lib/lib.es2022.error.d.ts:24

Inherited from

Error.cause

message

message: string;

Defined in: node_modules/.pnpm/typescript@6.0.3/node_modules/typescript/lib/lib.es5.d.ts:1075

Inherited from

Error.message

name

name: string;

Defined in: node_modules/.pnpm/typescript@6.0.3/node_modules/typescript/lib/lib.es5.d.ts:1074

Inherited from

Error.name

stack?

optional stack?: string;

Defined in: node_modules/.pnpm/typescript@6.0.3/node_modules/typescript/lib/lib.es5.d.ts:1076

Inherited from

Error.stack

stackTraceLimit

static stackTraceLimit: number;

Defined in: node_modules/.pnpm/@types+node@25.9.8/node_modules/@types/node/globals.d.ts:67

The Error.stackTraceLimit property specifies the number of stack frames collected by a stack trace (whether generated by new Error().stack or Error.captureStackTrace(obj)).

The default value is 10 but may be set to any valid JavaScript number. Changes will affect any stack trace captured after the value has been changed.

If set to a non-number value, or set to a negative number, stack traces will not capture any frames.

Inherited from

Error.stackTraceLimit

Methods

captureStackTrace()

static captureStackTrace(targetObject, constructorOpt?): void;

Defined in: node_modules/.pnpm/@types+node@25.9.8/node_modules/@types/node/globals.d.ts:51

Creates a .stack property on targetObject, which when accessed returns a string representing the location in the code at which Error.captureStackTrace() was called.

const myObject = {};
Error.captureStackTrace(myObject);
myObject.stack;  // Similar to `new Error().stack`

The first line of the trace will be prefixed with ${myObject.name}: ${myObject.message}.

The optional constructorOpt argument accepts a function. If given, all frames above constructorOpt, including constructorOpt, will be omitted from the generated stack trace.

The constructorOpt argument is useful for hiding implementation details of error generation from the user. For instance:

function a() {
  b();
}

function b() {
  c();
}

function c() {
  // Create an error without stack trace to avoid calculating the stack trace twice.
  const { stackTraceLimit } = Error;
  Error.stackTraceLimit = 0;
  const error = new Error();
  Error.stackTraceLimit = stackTraceLimit;

  // Capture the stack trace above function b
  Error.captureStackTrace(error, b); // Neither function c, nor b is included in the stack trace
  throw error;
}

a();

Parameters

targetObject

object

constructorOpt?

Function

Returns

void

Inherited from

Error.captureStackTrace

prepareStackTrace()

static prepareStackTrace(err, stackTraces): any;

Defined in: node_modules/.pnpm/@types+node@25.9.8/node_modules/@types/node/globals.d.ts:55

Parameters

err

Error

stackTraces

CallSite[]

Returns

any

See

https://v8.dev/docs/stack-trace-api#customizing-stack-traces

Inherited from

Error.prepareStackTrace

OWVoting

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:71

Opinion-Wise higher-order voting implementation. Uses Bayesian aggregation with correlation awareness.

Extended by

HigherOrderVotingStrategy

Constructors

Constructor

new OWVoting(options?): OWVoting;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:77

Parameters

options?

OWVotingOptions = {}

Returns

OWVoting

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning";

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:72

Implementation of

IVotingStrategy.algorithm

Methods

aggregate()

aggregate(votes, tracker?): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:199

Full pipeline: estimate correlation, compute result.

Parameters

votes

tracker?

Returns

IHigherOrderVoting.aggregate

Implementation of

aggregateWithCorrelation()

aggregateWithCorrelation(votes, correlationMatrix): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:92

Aggregate votes using Bayesian correlation-aware method.

Parameters

votes

correlationMatrix

Returns

IHigherOrderVoting.aggregateWithCorrelation

Implementation of

calculateOutcome()

calculateOutcome(votes, _weights?): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:87

IVotingStrategy implementation for integration with ConsensusEngine.

Parameters

votes

_weights?

Map<string, number>

Returns

IVotingStrategy.calculateOutcome

Implementation of

computeISP()

computeISP(votes, independentSubsets): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:152

Compute result using Independent Subset Partition method.

Parameters

votes

independentSubsets

readonly IndependentSubset[]

Returns

IHigherOrderVoting.computeISP

Implementation of

estimateCorrelation()

estimateCorrelation(tracker): CorrelationMatrix;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:148

Estimate correlation matrix from voting history.

Parameters

tracker

Returns

IHigherOrderVoting.estimateCorrelation

Implementation of

getConfig()

getConfig(): HigherOrderVotingConfig;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:251

Get the current configuration.

Returns

HigherOrderVotingConfig

Implementation of

IHigherOrderVoting.getConfig

ProofOfLearningStrategy

Defined in: packages/nexus-agents/src/consensus/strategies.ts:265

Proof-of-learning weighted voting strategy. Agents with better track records have more voting power.

Extends

BaseVotingStrategy

Constructors

Constructor

new ProofOfLearningStrategy(): ProofOfLearningStrategy;

Returns

ProofOfLearningStrategy

Inherited from

BaseVotingStrategy.constructor

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = 'proof_of_learning';

Defined in: packages/nexus-agents/src/consensus/strategies.ts:266

Overrides

BaseVotingStrategy.algorithm

Methods

calculateOutcome()

calculateOutcome(votes, weights?): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:268

Parameters

votes

weights?

Map<string, number>

Returns

Overrides

BaseVotingStrategy.calculateOutcome

countVotes()

protected countVotes(votes): VoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:80

Count votes by decision type.

Parameters

votes

Returns

Inherited from

BaseVotingStrategy.countVotes

countWeightedVotes()

protected countWeightedVotes(votes, weights): WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:105

Calculate weighted vote counts using agent performance weights.

Parameters

votes

weights

Map<string, number>

Returns

Inherited from

BaseVotingStrategy.countWeightedVotes

SimpleMajorityStrategy

Defined in: packages/nexus-agents/src/consensus/strategies.ts:138

Simple majority voting strategy (>50% approval).

Extends

BaseVotingStrategy

Constructors

Constructor

new SimpleMajorityStrategy(): SimpleMajorityStrategy;

Returns

SimpleMajorityStrategy

Inherited from

BaseVotingStrategy.constructor

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = 'simple_majority';

Defined in: packages/nexus-agents/src/consensus/strategies.ts:139

Overrides

BaseVotingStrategy.algorithm

Methods

calculateOutcome()

calculateOutcome(votes): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:141

Parameters

votes

Returns

Overrides

BaseVotingStrategy.calculateOutcome

countVotes()

protected countVotes(votes): VoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:80

Count votes by decision type.

Parameters

votes

Returns

Inherited from

BaseVotingStrategy.countVotes

countWeightedVotes()

protected countWeightedVotes(votes, weights): WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:105

Calculate weighted vote counts using agent performance weights.

Parameters

votes

weights

Map<string, number>

Returns

Inherited from

BaseVotingStrategy.countWeightedVotes

SupermajorityStrategy

Defined in: packages/nexus-agents/src/consensus/strategies.ts:176

Supermajority voting strategy (at least 2/3 approval).

Extends

BaseVotingStrategy

Constructors

Constructor

new SupermajorityStrategy(): SupermajorityStrategy;

Returns

SupermajorityStrategy

Inherited from

BaseVotingStrategy.constructor

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = 'supermajority';

Defined in: packages/nexus-agents/src/consensus/strategies.ts:177

Overrides

BaseVotingStrategy.algorithm

Methods

calculateOutcome()

calculateOutcome(votes): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:179

Parameters

votes

Returns

Overrides

BaseVotingStrategy.calculateOutcome

countVotes()

protected countVotes(votes): VoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:80

Count votes by decision type.

Parameters

votes

Returns

Inherited from

BaseVotingStrategy.countVotes

countWeightedVotes()

protected countWeightedVotes(votes, weights): WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:105

Calculate weighted vote counts using agent performance weights.

Parameters

votes

weights

Map<string, number>

Returns

Inherited from

BaseVotingStrategy.countWeightedVotes

UnanimousStrategy

Defined in: packages/nexus-agents/src/consensus/strategies.ts:215

Unanimous voting strategy (100% approval required).

Extends

BaseVotingStrategy

Constructors

Constructor

new UnanimousStrategy(): UnanimousStrategy;

Returns

UnanimousStrategy

Inherited from

BaseVotingStrategy.constructor

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = 'unanimous';

Defined in: packages/nexus-agents/src/consensus/strategies.ts:216

Overrides

BaseVotingStrategy.algorithm

Methods

calculateOutcome()

calculateOutcome(votes): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:218

Parameters

votes

Returns

Overrides

BaseVotingStrategy.calculateOutcome

countVotes()

protected countVotes(votes): VoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:80

Count votes by decision type.

Parameters

votes

Returns

Inherited from

BaseVotingStrategy.countVotes

countWeightedVotes()

protected countWeightedVotes(votes, weights): WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:105

Calculate weighted vote counts using agent performance weights.

Parameters

votes

weights

Map<string, number>

Returns

IVotingProtocol.createSession

Inherited from

BaseVotingStrategy.countWeightedVotes

VotingProtocol

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:49

Multi-round voting protocol for code review.

Implements

IVotingProtocol

Constructors

Constructor

new VotingProtocol(customLogger?): VotingProtocol;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:53

Parameters

customLogger?

ILogger

Returns

VotingProtocol

Methods

createSession()

createSession(
   topic, 
   committee, 
   config?
): VotingSession;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:60

Create a new voting session with a committee.

Parameters

topic

string

committee

string[]

config?

Partial<VotingProtocolConfig>

Returns

VotingSession

Implementation of

detectSycophancy()

detectSycophancy(sessionId): SycophancyReport;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:345

Detect sycophancy patterns.

Parameters

sessionId

string

Returns

SycophancyReport

Implementation of

IVotingProtocol.detectSycophancy

getResult()

getResult(sessionId): Promise<VotingProtocolResult | null>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:298

Get the final result.

Parameters

sessionId

string

Returns

Promise<VotingProtocolResult | null>

Implementation of

IVotingProtocol.getResult

getSession()

getSession(sessionId): VotingSession | undefined;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:353

Get the current session state.

Parameters

sessionId

string

Returns

VotingSession | undefined

Implementation of

IVotingProtocol.getSession

startAnalysisRound()

startAnalysisRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:99

Start the analysis round (Round 1).

Parameters

sessionId

string

Returns

Promise<VotingRound>

Implementation of

IVotingProtocol.startAnalysisRound

startConsensusRound()

startConsensusRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:238

Start the consensus round (Round 3).

Parameters

sessionId

string

Returns

Promise<VotingRound>

Implementation of

IVotingProtocol.startConsensusRound

startDeliberationRound()

startDeliberationRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:153

Start the deliberation round (Round 2).

Parameters

sessionId

string

Returns

Promise<VotingRound>

Implementation of

IVotingProtocol.startDeliberationRound

submitFinalVote()

submitFinalVote(
   sessionId, 
   agentId, 
   vote
): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:263

Submit final vote during consensus.

Parameters

sessionId

string

agentId

string

vote

conditions?

string[] = ...

confidence

number = ...

decision

"approve" | "reject" | "abstain" = VoteDecisionSchema

findings?

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

string = ...

rejectionCategories?

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

string = ...

timestamp?

string = ...

Returns

Promise<void>

Implementation of

IVotingProtocol.submitFinalVote

submitFindings()

submitFindings(
   sessionId, 
   agentId, 
   findings
): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:118

Submit findings from an agent during analysis.

Parameters

sessionId

string

agentId

string

findings

{ agentId: string; category: | "security" | "documentation" | "performance" | "design" | "other" | "bug" | "style"; confidence: number; description: string; location?: string; severity: "critical" | "major" | "minor" | "suggestion"; suggestion?: string; timestamp?: string; }[]

Returns

Promise<void>

Implementation of

IVotingProtocol.submitFindings

voteOnFinding()

voteOnFinding(sessionId, vote): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/voting-protocol.ts:199

Vote on findings during deliberation.

Parameters

sessionId

string

vote

agentId

string = ...

agree

boolean = ...

amendedSeverity?

"critical" | "major" | "minor" | "suggestion" = ...

findingId

string = ...

reasoning?

string = ...

Returns

Promise<void>

Implementation of

IVotingProtocol.voteOnFinding

VotingStrategyFactory

Defined in: packages/nexus-agents/src/consensus/strategies.ts:352

Factory for creating voting strategies.

Constructors

Constructor

new VotingStrategyFactory(): VotingStrategyFactory;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:355

Returns

VotingStrategyFactory

Methods

getAvailableAlgorithms()

getAvailableAlgorithms(): (
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning")[];

Defined in: packages/nexus-agents/src/consensus/strategies.ts:387

Get all available algorithm types.

Returns

getStrategy()

getStrategy(algorithm): IVotingStrategy;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:369

Get a voting strategy by algorithm type.

Parameters

algorithm

Returns

IVotingStrategy

registerStrategy()

registerStrategy(strategy): void;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:380

Parameters

strategy

IVotingStrategy

Returns

void

WeightedVoting

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:52

Weighted Byzantine voting implementation. Implements CP-WBFT pattern for fault-tolerant multi-agent consensus.

Implements

IWeightedVoting

Constructors

Constructor

new WeightedVoting(options?): WeightedVoting;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:58

Parameters

options?

WeightedVotingOptions = {}

Returns

WeightedVoting

Methods

calculateWeight()

calculateWeight(agentId): number;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:68

Calculate vote weight for an agent

Parameters

agentId

string

Returns

number

Implementation of

IWeightedVoting.calculateWeight

canVote()

canVote(agentId): boolean;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:185

Check if agent can vote

Parameters

agentId

string

Returns

boolean

Implementation of

IWeightedVoting.canVote

flagByzantine()

flagByzantine(agentId, reason): void;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:152

Flag agent for Byzantine behavior

Parameters

agentId

string

reason

string

Returns

void

Implementation of

IWeightedVoting.flagByzantine

getAgentRecord()

getAgentRecord(agentId): WeightedAgentRecord | undefined;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:146

Get agent performance record

Parameters

agentId

string

Returns

WeightedAgentRecord | undefined

Implementation of

IWeightedVoting.getAgentRecord

getAllRecords()

getAllRecords(): readonly WeightedAgentRecord[];

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:181

Get all agent records

Returns

readonly WeightedAgentRecord[]

Implementation of

IWeightedVoting.getAllRecords

recalibrateWeights()

recalibrateWeights(): void;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:191

Recalibrate all weights based on global performance

Returns

void

Implementation of

IWeightedVoting.recalibrateWeights

registerAgent()

registerAgent(agentId): void;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:136

Parameters

agentId

string

Returns

void

Implementation of

IWeightedVoting.registerAgent

updatePerformance()

updatePerformance(agentId, outcome): void;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:74

Update agent performance based on task outcome

Parameters

agentId

string

outcome

"unknown" | "success" | "failure" | "partial"

Returns

void

Implementation of

IWeightedVoting.updatePerformance

weightedConsensus()

weightedConsensus(votes): WeightedConsensusResult;

Defined in: packages/nexus-agents/src/consensus/weighted-voting.ts:109

Run weighted consensus on votes

Parameters

votes

Returns

WeightedConsensusResult

Implementation of

IWeightedVoting.weightedConsensus

Interfaces

AgentPerformance

Defined in: packages/nexus-agents/src/consensus/types-core.ts:235

Agent performance record for proof-of-learning.

Properties

agentId

agentId: string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:236

correctVotes

correctVotes: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:238

lastUpdated

lastUpdated: string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:240

successRate

successRate: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:239

totalVotes

totalVotes: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:237

AgentVoteResult

Defined in: packages/nexus-agents/src/cli/vote-types.ts:178

Individual agent vote with metadata.

Properties

assignedCli?

readonly optional assignedCli?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:262

The CLI the round-robin or NEXUS_VOTER_MODEL_<ROLE> pin chose for this seat (#6115), as a bare name (claude, not cli-claude). Unlike pinnedModel it is known BEFORE detection, so it survives the pending-detection placeholder and says where a seat was meant to answer. Compare with cli to see where it did. Absent on results built outside the panel launcher (simulation, direct executeAgentVote calls).

cacheCreationInputTokens?

readonly optional cacheCreationInputTokens?: number;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:316

Input tokens spent writing the cache, when reported (#4435). Bills at roughly 1.25x the uncached rate — the opposite end from a cache read.

cachedInputTokens?

readonly optional cachedInputTokens?: number;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:311

Input tokens read from an existing prompt cache, when the adapter reported them (#4435). Separate from inputTokens because cache reads bill at roughly a tenth of the uncached rate.

cli?

readonly optional cli?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:201

CLI that executed this vote (for adaptive routing feedback).

error?

readonly optional error?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:318

Error message if vote fell back to simulation or encountered an error

fallback?

readonly optional fallback?: SeatFallback;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:281

Present only when the seat answered on a different CLI or model than assigned (#6115). Three consecutive 7-seat panels ran every seat on one model because claude was out of credits and codex could not spawn, and nothing in the result said so — a single-model panel is a weaker independence claim than the assignment, and the tally read identically.

gatewayArm?

readonly optional gatewayArm?: `api:${string}`;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:273

The api:<endpoint> gateway arm this seat answered through, when it answered through an in-process gateway adapter (#4392 increment 2, step 4). Absent on a CLI-subprocess seat and on results built outside the panel launcher. The decision-cost rollup keys on it: a gateway seat is priced by the arm’s NEXUS_GATEWAY_COST declaration — UNKNOWN when undeclared — never by the model id’s vendor list price, which is what a claude-* id served by a gateway used to record. cli stays the adapter’s providerId (openai), which cannot say this.

inputTokens?

readonly optional inputTokens?: number;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:300

Input tokens the adapter reported for this voter’s LLM call, when known (#3910). Propagated from CompletionResponse.usage so per-decision cost aggregation resolves from unmeasured to MEASURED. Absent for error/simulation votes that never reached a model, or for adapters that do not report usage (CLI subscriptions) — those stay honestly unmeasured.

model?

readonly optional model?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:238

Model id that executed this vote, when known (e.g. ‘claude-sonnet’). Carried so per-decision cost aggregation can attribute spend per model (#3855). Absent for error/simulation votes that never reached a model.

outputTokens?

readonly optional outputTokens?: number;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:305

Output tokens the adapter reported for this voter’s LLM call, when known (#3910). See AgentVoteResult.inputTokens.

pinnedModel?

readonly optional pinnedModel?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:253

Model assigned to this role before execution. Unlike model, this stays on the primary assignment when router failover serves the vote elsewhere.

processingTimeMs

readonly processingTimeMs: number;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:181

retried?

readonly optional retried?: boolean;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:223

True when this vote came from the per-role retry of an errored seat (#5578). A first-attempt vote never carries it.

The panel launches once; a voter that errors is dropped, so under reduce_denominator its seat silently leaves the denominator and under absolute_quorum the whole vote voids and the caller replays all N voters for a single failure. Retrying just the errored roles recovers the seat for one extra call — and this flag is what makes the recovery visible instead of indistinguishable from a clean first attempt.

retriedFrom?

readonly optional retriedFrom?: RetriedFrom;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:232

What this seat was retried from (#6246): the first pass’s source and, when it had one, its error string. Present only on a seat the per-role retry REPLACED — never on a first-attempt result, and never on a seat whose retry failed again (that seat keeps its first attempt, unmarked). Orthogonal to retried: that flag says a recovery happened; this says what it recovered from.

role

readonly role: VoterRole;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:179

selectedOption?

readonly optional selectedOption?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:211

Which named option this voter chose, when the proposal declared options (#4452). Absent on an ordinary yes/no vote.

The approve/reject/abstain tally cannot express option choice: on a multi-option proposal every engaged voter returns approve, so a real 6-1 split records as unanimous. This is what makes the split recoverable without parsing free-text reasoning.

servedModel?

readonly optional servedModel?: string;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:248

Model id the adapter REPORTED serving this vote (CompletionResponse.model), when it reported one (#6660). Differs from model — the adapter’s construction-time id, i.e. the model REQUESTED — when the transport answered on another model: the claude adapter’s in-family substitution after an out-of-credits envelope (#6120) is the known case. The outcome row’s servedModel and its price are taken from this field, never from model. Absent for error/simulation votes and when the adapter reported no model.

source

readonly source: "error" | "llm" | "unverifiable" | "simulation";

Defined in: packages/nexus-agents/src/cli/vote-types.ts:193

Source of the vote:

‘llm’: Real LLM execution
‘simulation’: Fallback simulation (opt-in only)
‘error’: Error during execution (Issue #523)
‘unverifiable’: the seat answered but could not read the artifact (#6094). A DISTINCT value, not a flag on abstain, so no aggregation over vote.decision can fold it back into the abstain bucket. The vote.decision is always abstain — an unverifiable seat never carries approve/reject, whatever the model returned.

timing?

readonly optional timing?: {
  attempts: readonly SeatAttemptTiming[];
};

Defined in: packages/nexus-agents/src/cli/vote-types.ts:292

Per-attempt timing of this seat (#6103): how long each attempt QUEUED behind its CLI’s serialized lane (#3348) and how long it RAN, in launch order — the primary attempt first, then the cross-CLI fallback (#3587) when one was made. Absent on a seat that never reached the launcher (simulation, a direct executeAgentVote); an empty attempts list on a seat the launcher refused before any attempt (cancelled). Recorded so the panel wall-clock can be attributed to queueing versus model time before a fallback lane is designed; never folded into the vote record.

attempts

readonly attempts: readonly SeatAttemptTiming[];

unverifiableSignal?

readonly optional unverifiableSignal?: UnverifiableSignal;

Defined in: packages/nexus-agents/src/cli/vote-types.ts:199

Present only when source === 'unverifiable': which evidence classified the seat (#6094). Lets an auditor tell a structured stderr signal from the reasoning-text heuristic.

vote

readonly vote: {
  conditions?: string[];
  confidence: number;
  decision: "approve" | "reject" | "abstain";
  findings?: {
     claim: string;
     gate: {
        named_assertion: string;
        reread_cited_line: "failed" | "skipped" | "passed";
        ruled_out_language_non_issue: "failed" | "skipped" | "passed";
        traced_call_path: "failed" | "skipped" | "passed";
     };
     location: string;
     severity: "critical" | "high" | "low" | "medium";
     summary: string;
  }[];
  reasoning: string;
  rejectionCategories?: (
     | "YAGNI"
     | "DRY_VIOLATION"
     | "OVER_ENGINEERING"
     | "SCOPE_CREEP"
     | "SECURITY_RISK"
     | "MISALIGNED"
    | "INSUFFICIENT_EVIDENCE")[];
  selectedOption?: string;
  timestamp?: string;
};

Defined in: packages/nexus-agents/src/cli/vote-types.ts:180

conditions?

optional conditions?: string[];

confidence

confidence: number;

decision

decision: "approve" | "reject" | "abstain" = VoteDecisionSchema;

findings?

optional findings?: {
  claim: string;
  gate: {
     named_assertion: string;
     reread_cited_line: "failed" | "skipped" | "passed";
     ruled_out_language_non_issue: "failed" | "skipped" | "passed";
     traced_call_path: "failed" | "skipped" | "passed";
  };
  location: string;
  severity: "critical" | "high" | "low" | "medium";
  summary: string;
}[];

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

reasoning: string;

rejectionCategories?

optional rejectionCategories?: (
  | "YAGNI"
  | "DRY_VIOLATION"
  | "OVER_ENGINEERING"
  | "SCOPE_CREEP"
  | "SECURITY_RISK"
  | "MISALIGNED"
  | "INSUFFICIENT_EVIDENCE")[];

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

optional selectedOption?: string;

timestamp?

optional timestamp?: string;

CollectRealVotesOptions

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:276

Options for collecting votes from multiple agents.

Extends

VoterAgentOptions

Properties

adapter?

readonly optional adapter?: IModelAdapter;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:113

Model adapter to use (auto-selected if not provided)

Inherited from

VoterAgentOptions.adapter

allowSimulation?

readonly optional allowSimulation?: boolean;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:119

Whether to allow simulation fallback (default: false per Issue #280)

Inherited from

VoterAgentOptions.allowSimulation

declaredOptions?

readonly optional declaredOptions?: readonly string[];

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:296

Named alternatives for a multi-option proposal (#4472); each voter picks one.

erroredRoleBackoffMs?

readonly optional erroredRoleBackoffMs?: number;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:320

Delay before the per-role retry of errored voters (#5578). Defaults to DEFAULT_ERRORED_ROLE_BACKOFF_MS; 0 disables the wait, which is what tests want. The retry itself is not optional — the panel applies it under every error policy, per the #5578 design panel (option b, 6 of 6).

gatewayAdapters?

readonly optional gatewayAdapters?: readonly IModelAdapter[];

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:294

In-process gateway model adapters (#4040) — one per discovered gateway model. When provided (and no explicit adapter override), voters route through these HTTP adapters instead of shelling out to CLIs: roles are round-robined across them for per-role model diversity, the API key stays in-process (no subprocess env-forwarding), and the nested-spawn class (#4033) cannot occur. Empty/omitted ⇒ the CLI round-robin path is used (unchanged).

interAgentDelayMs?

readonly optional interAgentDelayMs?: number;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:121

Delay between launching each agent vote to prevent rate limiting (default: 1000ms). Set to 0 to disable.

Inherited from

VoterAgentOptions.interAgentDelayMs

logger?

readonly optional logger?: ILogger;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:111

Logger instance

Inherited from

VoterAgentOptions.logger

maxRetries?

readonly optional maxRetries?: number;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:117

Maximum retries per vote (default: 2)

Inherited from

VoterAgentOptions.maxRetries

onVoteCollected?

readonly optional onVoteCollected?: (vote) => void;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:313

Per-seat progress callback (#6162): invoked with each seat’s result as it settles, on every pass (first, fallback, retry). The async-job heartbeat for vote bodies; absent changes nothing.

Parameters

vote

AgentVoteResult

Returns

void

project?

readonly optional project?: string;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:301

The project the panel is judging, already resolved and validated by the caller (#6110). Reaches every seat’s system prompt. Absent ⇒ nexus-agents.

proposal

readonly proposal: string;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:283

Proposal text

roleAdapters?

readonly optional roleAdapters?: ReadonlyMap<VoterRole, IModelAdapter>;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:327

Seats already assigned by assignPanelSeats (#6003): a caller that must know the panel’s models BEFORE the vote (the pr_review budget reads their windows) resolves the seats once and hands them back, so the panel runs on exactly what it was budgeted for. Ignored when adapter is set.

roles

readonly roles: readonly VoterRole[];

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:281

Voter roles to include

signal?

readonly optional signal?: AbortSignal;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:307

Cancellation for an in-flight panel. Stops LAUNCHING voters that have not started (#5393), aborts the adapter call of every seat already in flight (#6729), and skips the errored-seat retry pass. Absent changes nothing.

simulate?

readonly optional simulate?: boolean;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:285

Use simulation mode (explicit opt-in only)

timeoutMs?

readonly optional timeoutMs?: number;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:115

Timeout per vote in milliseconds (default: 120000, override via NEXUS_VOTE_TIMEOUT_MS)

Inherited from

VoterAgentOptions.timeoutMs

workspace?

readonly optional workspace?: string;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:278

Repository checkout used by every panel seat.

workspaceSha?

readonly optional workspaceSha?: string;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:279

ConsensusEngineConfig

Defined in: packages/nexus-agents/src/consensus/types-core.ts:307

Consensus engine configuration.

Properties

defaultTimeout

defaultTimeout: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:308

enablePerformanceTracking

enablePerformanceTracking: boolean;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:311

incrementalQuorum?

optional incrementalQuorum?: IncrementalQuorumConfig;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:317

Incremental quorum configuration (Issue #1408)

maxActiveProposals

maxActiveProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:310

maxClosedProposals

maxClosedProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:313

Maximum number of closed proposals to retain. Oldest are evicted when exceeded. (Issue #549)

minVotersForQuorum

minVotersForQuorum: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:309

proposalCache?

optional proposalCache?: ProposalCacheConfig;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:315

Content-based proposal caching for determinism (Issue #589)

ConsensusMetrics

Defined in: packages/nexus-agents/src/consensus/types-core.ts:384

Consensus metrics for monitoring.

Properties

algorithmUsage

algorithmUsage: Record<ConsensusAlgorithm, number>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:391

approvedProposals

approvedProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:386

averageDurationMs

averageDurationMs: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:389

averageVotesPerProposal

averageVotesPerProposal: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:390

rejectedProposals

rejectedProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:387

timedOutProposals

timedOutProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:388

totalProposals

totalProposals: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:385

ConsensusResult

Defined in: packages/nexus-agents/src/consensus/types-core.ts:180

Result of a consensus decision.

Properties

approvalPercentage

approvalPercentage: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:196

closedAt

closedAt: string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:199

durationMs

durationMs: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:200

outcome

outcome: "timeout" | "closed" | "rejected" | "pending" | "voting" | "approved";

Defined in: packages/nexus-agents/src/consensus/types-core.ts:183

proposal

proposal: {
  algorithm:   | "simple_majority"
     | "supermajority"
     | "unanimous"
     | "higher_order"
     | "opinion_wise"
     | "proof_of_learning";
  createdAt?: string;
  description: string;
  id?: string;
  metadata?: Record<string, unknown>;
  requiredVoters?: string[];
  timeout?: number;
  title: string;
};

Defined in: packages/nexus-agents/src/consensus/types-core.ts:182

algorithm

algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = ConsensusAlgorithmSchema;

createdAt?

optional createdAt?: string;

description

description: string;

id?

optional id?: string;

metadata?

optional metadata?: Record<string, unknown>;

requiredVoters?

optional requiredVoters?: string[];

timeout?

optional timeout?: number;

title

title: string;

proposalId

proposalId: string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:181

quorumReached

quorumReached: boolean;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:197

startedAt

startedAt: string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:198

voteCounts

voteCounts: VoteCounts;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:185

votes

votes: Map<string, {
  conditions?: string[];
  confidence: number;
  decision: "approve" | "reject" | "abstain";
  findings?: {
     claim: string;
     gate: {
        named_assertion: string;
        reread_cited_line: "failed" | "skipped" | "passed";
        ruled_out_language_non_issue: "failed" | "skipped" | "passed";
        traced_call_path: "failed" | "skipped" | "passed";
     };
     location: string;
     severity: "critical" | "high" | "low" | "medium";
     summary: string;
  }[];
  reasoning: string;
  rejectionCategories?: (
     | "YAGNI"
     | "DRY_VIOLATION"
     | "OVER_ENGINEERING"
     | "SCOPE_CREEP"
     | "SECURITY_RISK"
     | "MISALIGNED"
    | "INSUFFICIENT_EVIDENCE")[];
  selectedOption?: string;
  timestamp?: string;
}>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:184

weightBasis?

optional weightBasis?: WeightBasis;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:195

What the weights in weightedCounts were derived from (#5117).

Carried on the RESULT, not just computed inside the strategy, because the result is what a reviewer reads. A tally reported as weighted when every weight was structurally 1.0 invites the reader to believe voter track record moved the number. Absent for strategies that do not weight at all.

weightedCounts?

optional weightedCounts?: WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:186

ConsolidatedFinding

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:154

A consolidated finding after deliberation.

Properties

agreementRatio

agreementRatio: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:162

description

description: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:158

id

id: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:155

location?

optional location?: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:159

originalFindings

originalFindings: {
  agentId: string;
  category:   | "security"
     | "documentation"
     | "performance"
     | "design"
     | "other"
     | "bug"
     | "style";
  confidence: number;
  description: string;
  location?: string;
  severity: "critical" | "major" | "minor" | "suggestion";
  suggestion?: string;
  timestamp?: string;
}[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:163

agentId

agentId: string;

confidence

confidence: number;

description

description: string;

location?

optional location?: string;

severity

severity: "critical" | "major" | "minor" | "suggestion";

suggestion?

optional suggestion?: string;

timestamp?

optional timestamp?: string;

severity

severity: "critical" | "major" | "minor" | "suggestion";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:157

suggestion?

optional suggestion?: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:160

supportingAgents

supportingAgents: string[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:161

CorrelationTrackerStats

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:290

Statistics about correlation tracking.

Properties

averageCorrelation

readonly averageCorrelation: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:298

Average correlation across all pairs

independentSubsetCount

readonly independentSubsetCount: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:300

Number of identified independent subsets

pairsWithSufficientData

readonly pairsWithSufficientData: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:302

Pairs with sufficient data for correlation calculation

totalAgents

readonly totalAgents: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:292

Total agents being tracked

totalObservations

readonly totalObservations: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:296

Total voting observations recorded

trackedPairs

readonly trackedPairs: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:294

Total agent pairs with correlation data

HigherOrderVotingConfig

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:185

Configuration for higher-order voting. Correlation aggregates are lifetime evidence; retained records are count-bounded by maxProposals FIFO and maxObservationsPerAgent, and active history is partitioned by each role’s pinned model. The legacy correlation lifetime keys are deprecated and ignored.

Properties

correlationMaxAgeMs

readonly correlationMaxAgeMs: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:193

Deprecated

Ignored since 8.x: correlation evidence is lifetime and partitioned by the role’s pinned model (#5555); nothing reads this value. Removed in the next major (#5564).

correlationThreshold

readonly correlationThreshold: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:189

Correlation threshold to consider agents correlated (default: 0.3)

fallbackToSimpleVoting

readonly fallbackToSimpleVoting: boolean;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:197

Whether to fall back to simple voting when correlation data insufficient

independenceThreshold

readonly independenceThreshold: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:195

Independence threshold for ISP grouping (default: 0.2)

maxObservationsPerAgent

readonly maxObservationsPerAgent: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:203

Maximum observations to store per agent before FIFO eviction (default: 1000)

maxProposals

readonly maxProposals: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:205

Maximum total proposals to track before evicting oldest (default: 5000)

maxTrackedPairs

readonly maxTrackedPairs: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:207

Maximum pairwise history entries before LRU eviction (default: 100)

minObservationsForCorrelation

readonly minObservationsForCorrelation: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:187

Minimum observations before using correlation data (default: 10)

observationDecayFactor

readonly observationDecayFactor: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:201

Deprecated

Ignored since 8.x: correlation evidence is lifetime and partitioned by the role’s pinned model (#5555); nothing reads this value. Removed in the next major (#5564).

HigherOrderVotingResult

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:247

Result of Bayesian aggregation with correlation awareness.

Properties

decision

readonly decision: "approve" | "reject" | "no_consensus";

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:249

Final decision

downweightedAgents

readonly downweightedAgents: readonly string[];

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:265

Agents whose votes were down-weighted due to correlation

effectiveVoteCount

readonly effectiveVoteCount: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:255

Effective number of independent votes

improvementOverBaseline

readonly improvementOverBaseline: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:261

Improvement over baseline majority voting (percentage points)

independentSubsets?

readonly optional independentSubsets?: readonly IndependentSubset[];

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:263

Independent subsets used (if ISP method)

method

readonly method: "simple" | "ow" | "isp";

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:259

Method used: ‘ow’ (opinion-wise), ‘isp’, or ‘simple’ (fallback)

posteriorApproval

readonly posteriorApproval: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:251

Posterior probability of approval

posteriorRejection

readonly posteriorRejection: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:253

Posterior probability of rejection

reasoning

readonly reasoning: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:267

Reasoning for the decision

usedCorrelationData

readonly usedCorrelationData: boolean;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:257

Whether correlation data was sufficient

IConsensusEngine

Defined in: packages/nexus-agents/src/consensus/engine.ts:67

Interface for the consensus engine.

Methods

close()

close(proposalId): Promise<Result<ConsensusResult, ConsensusError>>;

Defined in: packages/nexus-agents/src/consensus/engine.ts:71

Parameters

proposalId

string

Returns

Promise<Result<ConsensusResult, ConsensusError>>

getMetrics()

getMetrics(): ConsensusMetrics;

Defined in: packages/nexus-agents/src/consensus/engine.ts:72

Returns

ConsensusMetrics

getResult()

getResult(proposalId): Promise<Result<ConsensusResult, ConsensusError>>;

Defined in: packages/nexus-agents/src/consensus/engine.ts:70

Parameters

proposalId

string

Returns

Promise<Result<ConsensusResult, ConsensusError>>

propose()

propose(proposal): Promise<Result<string, ConsensusError>>;

Defined in: packages/nexus-agents/src/consensus/engine.ts:68

Parameters

proposal

algorithm

createdAt?

string = ...

description

string = ...

id?

string = ...

metadata?

Record<string, unknown> = ...

requiredVoters?

string[] = ...

timeout?

number = ...

title

string = ...

Returns

Promise<Result<string, ConsensusError>>

vote()

vote(
   proposalId, 
   agentId, 
   vote
): Promise<Result<void, ConsensusError>>;

Defined in: packages/nexus-agents/src/consensus/engine.ts:69

Parameters

proposalId

string

agentId

string

vote

conditions?

string[] = ...

confidence

number = ...

decision

"approve" | "reject" | "abstain" = VoteDecisionSchema

findings?

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

string = ...

rejectionCategories?

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

string = ...

timestamp?

string = ...

Returns

Promise<Result<void, ConsensusError>>

ICorrelationTracker

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:317

Interface for correlation tracking between agents.

Methods

clear()

clear(): void;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:370

Clear all recorded data.

Returns

void

computeCorrelationMatrix()

computeCorrelationMatrix(): CorrelationMatrix;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:344

Compute the full correlation matrix for all tracked agents.

Returns

CorrelationTrackerStats

getCorrelation()

getCorrelation(agentA, agentB): number | undefined;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:350

Get correlation between two specific agents. Returns undefined if insufficient data.

Parameters

agentA

string

agentB

string

Returns

number | undefined

getStats()

getStats(): CorrelationTrackerStats;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:365

Get statistics about the correlation tracker.

Returns

hasSufficientData()

hasSufficientData(agentIds): boolean;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:360

Check if there is sufficient correlation data for a set of agents.

Parameters

agentIds

readonly string[]

Returns

boolean

identifyIndependentSubsets()

identifyIndependentSubsets(): readonly IndependentSubset[];

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:355

Identify groups of agents that vote independently.

Returns

readonly IndependentSubset[]

recordProposalVotes()

recordProposalVotes(
   proposalId, 
   votes, 
   outcome, 
   context?
): void;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:334

Record votes from multiple agents for the same proposal.

Parameters

proposalId

string

votes

outcome

"rejected" | "approved"

context?

CorrelationRecordContext

Returns

void

recordVote()

recordVote(
   agentId, 
   vote, 
   outcome, 
   context?
): void;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:324

Record a vote and its outcome for correlation tracking.

Parameters

agentId

string

vote

conditions?

string[] = ...

confidence

number = ...

decision

"approve" | "reject" | "abstain" = VoteDecisionSchema

findings?

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

string = ...

rejectionCategories?

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

string = ...

timestamp?

string = ...

outcome

"rejected" | "approved"

context?

CorrelationRecordContext

Returns

void

setCurrentModelPins()?

optional setCurrentModelPins(modelPins): void;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:319

Select the active model partition for each pinned role.

Parameters

modelPins

ReadonlyMap<string, string>

Returns

void

IHigherOrderVoting

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:380

Interface for Opinion-Wise higher-order voting.

Methods

aggregate()

aggregate(votes, tracker?): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:410

Full pipeline: estimate correlation, compute result.

Parameters

votes

tracker?

Returns

aggregateWithCorrelation()

aggregateWithCorrelation(votes, correlationMatrix): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:384

Aggregate votes using Bayesian correlation-aware method.

Parameters

votes

correlationMatrix

Returns

computeISP()

computeISP(votes, independentSubsets): HigherOrderVotingResult;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:397

Compute result using Independent Subset Partition method.

Parameters

votes

independentSubsets

readonly IndependentSubset[]

Returns

estimateCorrelation()

estimateCorrelation(tracker): CorrelationMatrix;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:392

Estimate correlation matrix from voting history.

Parameters

tracker

Returns

HigherOrderVotingConfig

getConfig()

getConfig(): HigherOrderVotingConfig;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:418

Get the current configuration.

Returns

IndependentSubset

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:63

A subset of agents that vote independently of each other. Used in ISP (Independent Subset Partition) method.

Properties

agentIds

readonly agentIds: readonly string[];

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:67

Agent IDs in this independent subset

id

readonly id: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:65

Unique identifier for this subset

independenceScore

readonly independenceScore: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:76

Average internal independence score (lower = more independent).

Averaged over the pairs that HAVE a correlation, so read it together with pairCoverage: a subset whose pairs were measured at 0 and one whose pairs were never observed both score 0, and 0 is the score that earns the maximum posterior weight.

observationCount

readonly observationCount: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:88

Number of observations supporting this grouping

pairCoverage

readonly pairCoverage: {
  observed: number;
  total: number;
};

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:86

How many of the subset’s agent pairs actually had a correlation, out of how many exist.

observed < total means independenceScore is an average over a subset of the pairs — the unobserved ones contributed nothing rather than being represented as unknown. A singleton has total: 0: there is no pair to observe, so its score is not a measurement at all.

observed

readonly observed: number;

total

readonly total: number;

IVotingProtocol

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:182

Interface for the multi-round voting protocol. (Source: Issue #100, arXiv:2512.21352)

Methods

createSession()

createSession(
   topic, 
   committee, 
   config?
): VotingSession;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:184

Create a new voting session with a committee

Parameters

topic

string

committee

string[]

config?

Partial<VotingProtocolConfig>

Returns

VotingSession

detectSycophancy()

detectSycophancy(sessionId): SycophancyReport;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:212

Check for sycophancy in the current round

Parameters

sessionId

string

Returns

SycophancyReport

getResult()

getResult(sessionId): Promise<VotingProtocolResult | null>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:209

Get the final result (closes session if complete)

Parameters

sessionId

string

Returns

Promise<VotingProtocolResult | null>

getSession()

getSession(sessionId): VotingSession | undefined;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:215

Get the current session state

Parameters

sessionId

string

Returns

VotingSession | undefined

startAnalysisRound()

startAnalysisRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:191

Start the analysis round (Round 1)

Parameters

sessionId

string

Returns

Promise<VotingRound>

startConsensusRound()

startConsensusRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:203

Start the consensus round (Round 3)

Parameters

sessionId

string

Returns

Promise<VotingRound>

startDeliberationRound()

startDeliberationRound(sessionId): Promise<VotingRound>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:197

Start the deliberation round (Round 2)

Parameters

sessionId

string

Returns

Promise<VotingRound>

submitFinalVote()

submitFinalVote(
   sessionId, 
   agentId, 
   vote
): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:206

Submit final vote during consensus

Parameters

sessionId

string

agentId

string

vote

conditions?

string[] = ...

confidence

number = ...

decision

"approve" | "reject" | "abstain" = VoteDecisionSchema

findings?

Pre-verified PR-review findings (#2245 v4 follow-up). Optional; populated only when the voter emits the structured top-level array.

reasoning

string = ...

rejectionCategories?

Structured rejection categories for reject→refine→re-vote loops (Issue #1213).

selectedOption?

string = ...

timestamp?

string = ...

Returns

Promise<void>

submitFindings()

submitFindings(
   sessionId, 
   agentId, 
   findings
): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:194

Submit findings from an agent during analysis

Parameters

sessionId

string

agentId

string

findings

Returns

Promise<void>

voteOnFinding()

voteOnFinding(sessionId, vote): Promise<void>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:200

Vote on findings during deliberation

Parameters

sessionId

string

vote

agentId

string = ...

agree

boolean = ...

amendedSeverity?

"critical" | "major" | "minor" | "suggestion" = ...

findingId

string = ...

reasoning?

string = ...

Returns

Promise<void>

IVotingStrategy

Defined in: packages/nexus-agents/src/consensus/strategies.ts:24

Interface for voting strategy implementations.

Properties

algorithm

readonly algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning";

Defined in: packages/nexus-agents/src/consensus/strategies.ts:25

Methods

calculateOutcome()

calculateOutcome(votes, weights?): VotingOutcome;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:26

Parameters

votes

weights?

Map<string, number>

Returns

WeightedConsensusResult

IWeightedVoting

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:115

Interface for weighted Byzantine voting. (Source: Issue #103, arXiv:2511.10400 - CP-WBFT)

Methods

calculateWeight()

calculateWeight(agentId): number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:117

Calculate vote weight for an agent

Parameters

agentId

string

Returns

number

canVote()

canVote(agentId): boolean;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:138

Check if agent can vote

Parameters

agentId

string

Returns

boolean

flagByzantine()

flagByzantine(agentId, reason): void;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:132

Flag agent for Byzantine behavior

Parameters

agentId

string

reason

string

Returns

void

getAgentRecord()

getAgentRecord(agentId): WeightedAgentRecord | undefined;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:129

Get agent performance record

Parameters

agentId

string

Returns

WeightedAgentRecord | undefined

getAllRecords()

getAllRecords(): readonly WeightedAgentRecord[];

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:135

Get all agent records

Returns

readonly WeightedAgentRecord[]

recalibrateWeights()

recalibrateWeights(): void;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:141

Recalibrate all weights based on global performance

Returns

void

registerAgent()

registerAgent(agentId): void;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:126

Parameters

agentId

string

Returns

void

updatePerformance()

updatePerformance(agentId, outcome): void;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:120

Update agent performance based on task outcome

Parameters

agentId

string

outcome

"unknown" | "success" | "failure" | "partial"

Returns

void

weightedConsensus()

weightedConsensus(votes): WeightedConsensusResult;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:123

Run weighted consensus on votes

Parameters

votes

Returns

OWVotingOptions

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:48

Options for creating OWVoting instance.

Properties

algorithm?

readonly optional algorithm?: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning";

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:55

Algorithm label this instance reports (#3168). Defaults to simple_majority for backward compatibility; HigherOrderVotingStrategy sets opinion_wise. Keeps the label consistent whether constructed directly or via a factory.

config?

readonly optional config?: Partial<HigherOrderVotingConfig>;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:49

tracker?

readonly optional tracker?: ICorrelationTracker;

Defined in: packages/nexus-agents/src/consensus/higher-order-voting.ts:64

Correlation tracker this instance uses when OWVoting.aggregate is called WITHOUT a per-call tracker (#3173). Injecting here lets higher-order voting be reused as a building block (autonomous agents, test harnesses, custom pipelines) without coupling to the MCP tool’s process-wide singleton or threading the tracker through every call. Omit it to keep the singleton model — the MCP consensus path injects/passes its persistent tracker.

PairwiseVotingHistory

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:150

Aggregated voting history for a pair of agents.

Properties

agreements

readonly agreements: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:156

Number of times both agents agreed

correlation

readonly correlation: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:160

Computed correlation coefficient

disagreements

readonly disagreements: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:158

Number of times agents disagreed

jointObservations

readonly jointObservations: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:154

Number of proposals where both agents voted

lastUpdated

readonly lastUpdated: Date;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:162

Last update timestamp

pairKey

readonly pairKey: `${string}:${string}`;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:152

Agent pair key

ProposalState

Defined in: packages/nexus-agents/src/consensus/types-core.ts:363

Internal proposal state managed by the engine.

Properties

expansionInFlight?

optional expansionInFlight?: boolean;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:378

True while a quorum expansion is awaiting its callback for this proposal. Concurrent vote() calls check this to avoid double- expanding across the await gap (Issue #2861). Per-proposal so independent proposals never block each other.

expansionRounds?

optional expansionRounds?: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:371

Number of incremental quorum expansions applied (Issue #1408).

proposal

proposal: {
  algorithm:   | "simple_majority"
     | "supermajority"
     | "unanimous"
     | "higher_order"
     | "opinion_wise"
     | "proof_of_learning";
  createdAt?: string;
  description: string;
  id?: string;
  metadata?: Record<string, unknown>;
  requiredVoters?: string[];
  timeout?: number;
  title: string;
};

Defined in: packages/nexus-agents/src/consensus/types-core.ts:364

algorithm

algorithm: 
  | "simple_majority"
  | "supermajority"
  | "unanimous"
  | "higher_order"
  | "opinion_wise"
  | "proof_of_learning" = ConsensusAlgorithmSchema;

createdAt?

optional createdAt?: string;

description

description: string;

id?

optional id?: string;

metadata?

optional metadata?: Record<string, unknown>;

requiredVoters?

optional requiredVoters?: string[];

timeout?

optional timeout?: number;

title

title: string;

startedAt

startedAt: Date;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:368

status

status: "timeout" | "closed" | "rejected" | "pending" | "voting" | "approved";

Defined in: packages/nexus-agents/src/consensus/types-core.ts:365

timeoutId?

optional timeoutId?: Timeout;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:369

votes

votes: Map<string, {
  conditions?: string[];
  confidence: number;
  decision: "approve" | "reject" | "abstain";
  findings?: {
     claim: string;
     gate: {
        named_assertion: string;
        reread_cited_line: "failed" | "skipped" | "passed";
        ruled_out_language_non_issue: "failed" | "skipped" | "passed";
        traced_call_path: "failed" | "skipped" | "passed";
     };
     location: string;
     severity: "critical" | "high" | "low" | "medium";
     summary: string;
  }[];
  reasoning: string;
  rejectionCategories?: (
     | "YAGNI"
     | "DRY_VIOLATION"
     | "OVER_ENGINEERING"
     | "SCOPE_CREEP"
     | "SECURITY_RISK"
     | "MISALIGNED"
    | "INSUFFICIENT_EVIDENCE")[];
  selectedOption?: string;
  timestamp?: string;
}>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:366

voteWeights

voteWeights: Map<string, number>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:367

RoundSummary

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:169

Summary of a single round.

Properties

agreementScore

agreementScore: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:174

durationMs

durationMs: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:175

findingsCount

findingsCount: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:172

phase

phase: "analysis" | "consensus" | "deliberation";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:171

roundNumber

roundNumber: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:170

votesCount

votesCount: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:173

SycophancyIndicator

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:232

Individual sycophancy indicator.

Properties

agents

agents: string[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:236

description

description: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:234

severity

severity: "high" | "low" | "medium";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:235

type

type: 
  | "premature_consensus"
  | "opinion_convergence"
  | "confidence_inflation"
  | "echo_chamber";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:233

SycophancyReport

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:221

Report from sycophancy detection.

Properties

affectedAgents

affectedAgents: string[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:225

confidenceScore

confidenceScore: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:223

detected

detected: boolean;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:222

indicators

indicators: SycophancyIndicator[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:224

recommendation

recommendation: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:226

VoteCounts

Defined in: packages/nexus-agents/src/consensus/types-core.ts:139

Vote counts summary.

Properties

abstain

abstain: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:142

approve

approve: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:140

reject

reject: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:141

total

total: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:143

VotingObservation

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:109

Record of a single voting observation for correlation tracking.

Properties

agentId

readonly agentId: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:113

Agent who cast the vote

alignedWithOutcome

readonly alignedWithOutcome: boolean;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:119

Whether the vote aligned with the final outcome

confidence

readonly confidence: number;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:117

Confidence level (0-1)

decision

readonly decision: "approve" | "reject" | "abstain";

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:115

The vote decision

modelKey?

readonly optional modelKey?: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:123

Pinned model identity used to select the correlation partition.

observedModel?

readonly optional observedModel?: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:125

Model that actually answered, retained for provenance only.

proposalId

readonly proposalId: string;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:111

Unique proposal ID

timestamp

readonly timestamp: Date;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:121

Timestamp of the vote

VotingOutcome

Defined in: packages/nexus-agents/src/consensus/strategies.ts:32

Result of a voting strategy calculation.

Properties

approvalPercentage

approvalPercentage: number;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:34

approved

approved: boolean;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:33

reason

reason: string;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:44

voteCounts

voteCounts: VoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:35

weightBasis?

optional weightBasis?: WeightBasis;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:43

Present on weighted strategies only. Absent means the strategy does not weight at all (simple majority, supermajority, unanimous) — which is different from 'unweighted', meaning a weighted strategy ran with nothing to weight by.

weightedCounts?

optional weightedCounts?: WeightedVoteCounts;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:36

VotingProtocolConfig

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:86

Configuration for the voting protocol.

Properties

agreementThreshold

agreementThreshold: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:94

Minimum agreement threshold (default: 2/3, approximately 0.667)

committeeSize

committeeSize: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:88

Number of agents in the committee (default: 3)

enableAntiSycophancy

enableAntiSycophancy: boolean;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:96

Enable anti-sycophancy detection (default: true)

maxRounds

maxRounds: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:90

Maximum rounds before forcing decision (default: 3)

roundTimeoutMs

roundTimeoutMs: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:92

Timeout per round in milliseconds (default: multi-llm-panel guard, 900000)

sycophancyThreshold

sycophancyThreshold: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:98

Similarity threshold for sycophancy detection (default: 0.8)

VotingProtocolResult

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:139

Final result of a voting protocol session.

Properties

agreementScore

agreementScore: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:145

consolidatedFindings

consolidatedFindings: ConsolidatedFinding[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:143

outcome

outcome: "rejected" | "approved" | "needs_revision" | "no_consensus";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:142

participatingAgents

participatingAgents: string[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:148

roundSummaries

roundSummaries: RoundSummary[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:144

sessionId

sessionId: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:140

sycophancyDetected

sycophancyDetected: boolean;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:146

topic

topic: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:141

totalDurationMs

totalDurationMs: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:147

VotingRound

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:71

A single voting round in the protocol.

Properties

completedAt?

optional completedAt?: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:79

finalVotes

finalVotes: Map<string, {
  conditions?: string[];
  confidence: number;
  decision: "approve" | "reject" | "abstain";
  findings?: {
     claim: string;
     gate: {
        named_assertion: string;
        reread_cited_line: "failed" | "skipped" | "passed";
        ruled_out_language_non_issue: "failed" | "skipped" | "passed";
        traced_call_path: "failed" | "skipped" | "passed";
     };
     location: string;
     severity: "critical" | "high" | "low" | "medium";
     summary: string;
  }[];
  reasoning: string;
  rejectionCategories?: (
     | "YAGNI"
     | "DRY_VIOLATION"
     | "OVER_ENGINEERING"
     | "SCOPE_CREEP"
     | "SECURITY_RISK"
     | "MISALIGNED"
    | "INSUFFICIENT_EVIDENCE")[];
  selectedOption?: string;
  timestamp?: string;
}>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:77

findings

findings: Map<string, {
  agentId: string;
  category:   | "security"
     | "documentation"
     | "performance"
     | "design"
     | "other"
     | "bug"
     | "style";
  confidence: number;
  description: string;
  location?: string;
  severity: "critical" | "major" | "minor" | "suggestion";
  suggestion?: string;
  timestamp?: string;
}>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:75

findingVotes

findingVotes: Map<string, {
  agentId: string;
  agree: boolean;
  amendedSeverity?: "critical" | "major" | "minor" | "suggestion";
  findingId: string;
  reasoning?: string;
}[]>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:76

id

id: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:72

phase

phase: "analysis" | "consensus" | "deliberation";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:73

roundNumber

roundNumber: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:80

startedAt

startedAt: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:78

status

status: "aborted" | "completed" | "pending" | "in_progress" | "awaiting_votes";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:74

VotingSession

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:123

Session state for a voting protocol instance.

Properties

committee

committee: string[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:126

completedAt?

optional completedAt?: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:132

config

config: VotingProtocolConfig;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:129

createdAt

createdAt: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:131

currentRound

currentRound: number;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:128

finalResult?

optional finalResult?: VotingProtocolResult;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:133

id

id: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:124

rounds

rounds: VotingRound[];

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:127

status

status: "aborted" | "completed" | "active";

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:130

topic

topic: string;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:125

WeightedAgentRecord

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:24

Extended agent performance with Byzantine detection.

Properties

agentId

readonly agentId: string;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:25

byzantineFlags

readonly byzantineFlags: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:33

createdAt

readonly createdAt: Date;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:35

failedTasks

readonly failedTasks: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:28

lastActive

readonly lastActive: Date;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:34

partialTasks

readonly partialTasks: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:29

successfulTasks

readonly successfulTasks: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:27

successRate

readonly successRate: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:30

totalTasks

readonly totalTasks: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:26

trustScore

readonly trustScore: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:32

weight

readonly weight: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:31

WeightedConsensusResult

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:55

Weighted consensus result.

Properties

byzantineDetected

readonly byzantineDetected: boolean;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:61

decision

readonly decision: "approve" | "reject" | "no_consensus";

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:56

participatingAgents

readonly participatingAgents: readonly string[];

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:62

quorumReached

readonly quorumReached: boolean;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:60

totalWeight

readonly totalWeight: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:59

weightBreakdown

readonly weightBreakdown: ReadonlyMap<string, number>;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:63

weightedApproval

readonly weightedApproval: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:57

weightedRejection

readonly weightedRejection: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:58

WeightedVoteCounts

Defined in: packages/nexus-agents/src/consensus/types-core.ts:170

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

NOTE: VoteDecision and VoteDecisionSchema are exported from agents.ts (collaboration module) NOTE: TaskOutcome and TaskOutcomeSchema are exported from learning.ts These are intentionally omitted to avoid duplicate export errors.

Properties

abstain

abstain: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:173

approve

approve: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:171

reject

reject: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:172

totalWeight

totalWeight: number;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:174

WeightedVotingConfig

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:69

Configuration for weighted Byzantine voting.

Properties

byzantineFlagThreshold

readonly byzantineFlagThreshold: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:81

Byzantine flag threshold for exclusion (default: 3)

initialWeight

readonly initialWeight: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:83

Initial weight for new agents (default: 0.5)

maxByzantineFraction

readonly maxByzantineFraction: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:73

Maximum Byzantine fault tolerance (default: 0.33)

minTrustScore

readonly minTrustScore: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:79

Trust score required to vote (default: 0.3)

minWeight

readonly minWeight: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:71

Minimum weight to participate in voting (default: 0.1)

quorumThreshold

readonly quorumThreshold: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:85

Quorum threshold for valid consensus (default: 2/3, approximately 0.667)

weightDecayFactor

readonly weightDecayFactor: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:75

Weight decay factor per failed task (default: 0.9)

weightRecoveryFactor

readonly weightRecoveryFactor: number;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:77

Weight recovery factor per successful task (default: 1.05)

WeightedVotingOptions

Defined in: packages/nexus-agents/src/consensus/weighted-voting-helpers.ts:36

Options for WeightedVoting constructor.

Properties

config?

optional config?: Partial<WeightedVotingConfig>;

Defined in: packages/nexus-agents/src/consensus/weighted-voting-helpers.ts:38

Configuration for voting thresholds and weights.

emitEvents?

optional emitEvents?: boolean;

Defined in: packages/nexus-agents/src/consensus/weighted-voting-helpers.ts:42

Whether to emit Byzantine detection events (default: true if eventBus provided).

eventBus?

optional eventBus?: ICollaborationEventBus;

Defined in: packages/nexus-agents/src/consensus/weighted-voting-helpers.ts:40

Optional event bus for Byzantine detection events (Issue #218).

Type Aliases

AgentFinding

type AgentFinding = z.infer<typeof AgentFindingSchema>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:54

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

AgentPairKey

type AgentPairKey = `${string}:${string}`;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:25

Pair of agent IDs for correlation tracking. Stored as “agentA:agentB” where agentA < agentB lexicographically.

ConsensusAlgorithm

type ConsensusAlgorithm = z.infer<typeof ConsensusAlgorithmSchema>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:28

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

CorrelationCoefficient

type CorrelationCoefficient = z.infer<typeof CorrelationCoefficientSchema>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:52

CorrelationMatrix

type CorrelationMatrix = Map<AgentPairKey, CorrelationCoefficient>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:57

Correlation matrix storing pairwise correlations between agents.

FindingVote

type FindingVote = z.infer<typeof FindingVoteSchema>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:66

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

Proposal

type Proposal = z.infer<typeof ProposalSchema>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:129

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

ProposalId

type ProposalId = string;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:134

Unique identifier for a proposal.

ProposalStatus

type ProposalStatus = z.infer<typeof ProposalStatusSchema>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:47

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

RejectionCategory

type RejectionCategory = z.infer<typeof RejectionCategorySchema>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:62

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

Vote

type Vote = z.infer<typeof VoteSchema>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:114

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

VotingRoundPhase

type VotingRoundPhase = z.infer<typeof VotingRoundPhaseSchema>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:27

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

VotingRoundStatus

type VotingRoundStatus = z.infer<typeof VotingRoundStatusSchema>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:39

Consensus exports - Voting protocols, consensus engine, and strategies Split from index.ts for file size compliance (Issue #285) Added to public API per Issue #351

Variables

AgentFindingSchema

const AgentFindingSchema: ZodObject<{
  agentId: ZodString;
  category: ZodEnum<{
     bug: "bug";
     design: "design";
     documentation: "documentation";
     other: "other";
     performance: "performance";
     security: "security";
     style: "style";
  }>;
  confidence: ZodNumber;
  description: ZodString;
  location: ZodOptional<ZodString>;
  severity: ZodEnum<{
     critical: "critical";
     major: "major";
     minor: "minor";
     suggestion: "suggestion";
  }>;
  suggestion: ZodOptional<ZodString>;
  timestamp: ZodOptional<ZodISODateTime>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:44

A finding submitted by an agent during analysis.

AgentPerformanceSchema

const AgentPerformanceSchema: ZodObject<{
  agentId: ZodString;
  correctVotes: ZodNumber;
  lastUpdated: ZodISODateTime;
  successRate: ZodNumber;
  totalVotes: ZodNumber;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:246

Agent performance schema.

ConsensusAlgorithmSchema

const ConsensusAlgorithmSchema: ZodEnum<{
  higher_order: "higher_order";
  opinion_wise: "opinion_wise";
  proof_of_learning: "proof_of_learning";
  simple_majority: "simple_majority";
  supermajority: "supermajority";
  unanimous: "unanimous";
}>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:20

Consensus algorithm types.

simple_majority: >50% of votes required
supermajority: >=67% of votes required
unanimous: 100% approval required
proof_of_learning: weighted voting based on agent performance
opinion_wise: higher-order voting with correlation awareness (Issue #333)
higher_order: alias for opinion_wise (Issue #514)

ConsensusEngineConfigSchema

const ConsensusEngineConfigSchema: ZodObject<{
  defaultTimeout: ZodDefault<ZodNumber>;
  enablePerformanceTracking: ZodDefault<ZodBoolean>;
  maxActiveProposals: ZodDefault<ZodNumber>;
  maxClosedProposals: ZodDefault<ZodNumber>;
  minVotersForQuorum: ZodDefault<ZodNumber>;
  proposalCache: ZodOptional<ZodObject<{
     enabled: ZodDefault<ZodBoolean>;
     maxEntries: ZodDefault<ZodNumber>;
     ttlMs: ZodDefault<ZodNumber>;
  }, $strip>>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:332

Consensus engine configuration schema.

ConsensusMetricsSchema

const ConsensusMetricsSchema: ZodObject<{
  algorithmUsage: ZodRecord<ZodEnum<{
     higher_order: "higher_order";
     opinion_wise: "opinion_wise";
     proof_of_learning: "proof_of_learning";
     simple_majority: "simple_majority";
     supermajority: "supermajority";
     unanimous: "unanimous";
  }>, ZodNumber>;
  approvedProposals: ZodNumber;
  averageDurationMs: ZodNumber;
  averageVotesPerProposal: ZodNumber;
  rejectedProposals: ZodNumber;
  timedOutProposals: ZodNumber;
  totalProposals: ZodNumber;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:397

Consensus metrics schema.

ConsensusResultSchema

const ConsensusResultSchema: ZodObject<{
  approvalPercentage: ZodNumber;
  closedAt: ZodISODateTime;
  durationMs: ZodNumber;
  outcome: ZodEnum<{
     approved: "approved";
     closed: "closed";
     pending: "pending";
     rejected: "rejected";
     timeout: "timeout";
     voting: "voting";
  }>;
  proposal: ZodObject<{
     algorithm: ZodEnum<{
        higher_order: "higher_order";
        opinion_wise: "opinion_wise";
        proof_of_learning: "proof_of_learning";
        simple_majority: "simple_majority";
        supermajority: "supermajority";
        unanimous: "unanimous";
     }>;
     createdAt: ZodOptional<ZodISODateTime>;
     description: ZodString;
     id: ZodOptional<ZodString>;
     metadata: ZodOptional<ZodRecord<ZodString, ZodUnknown>>;
     requiredVoters: ZodOptional<ZodArray<ZodString>>;
     timeout: ZodOptional<ZodNumber>;
     title: ZodString;
  }, $strip>;
  proposalId: ZodString;
  quorumReached: ZodBoolean;
  startedAt: ZodISODateTime;
  voteCounts: ZodObject<{
     abstain: ZodNumber;
     approve: ZodNumber;
     reject: ZodNumber;
     total: ZodNumber;
  }, $strip>;
  votes: ZodMap<ZodString, ZodObject<{
     conditions: ZodOptional<ZodArray<ZodString>>;
     confidence: ZodNumber;
     decision: ZodEnum<{
        abstain: "abstain";
        approve: "approve";
        reject: "reject";
     }>;
     findings: ZodOptional<ZodArray<ZodObject<{
        claim: ZodString;
        gate: ZodObject<{
           named_assertion: ...;
           reread_cited_line: ...;
           ruled_out_language_non_issue: ...;
           traced_call_path: ...;
        }, $strip>;
        location: ZodString;
        severity: ZodDefault<ZodEnum<...>>;
        summary: ZodString;
     }, $strip>>>;
     reasoning: ZodString;
     rejectionCategories: ZodOptional<ZodArray<ZodEnum<{
        DRY_VIOLATION: "DRY_VIOLATION";
        INSUFFICIENT_EVIDENCE: "INSUFFICIENT_EVIDENCE";
        MISALIGNED: "MISALIGNED";
        OVER_ENGINEERING: "OVER_ENGINEERING";
        SCOPE_CREEP: "SCOPE_CREEP";
        SECURITY_RISK: "SECURITY_RISK";
        YAGNI: "YAGNI";
     }>>>;
     selectedOption: ZodOptional<ZodString>;
     timestamp: ZodOptional<ZodISODateTime>;
  }, $strip>>;
  weightedCounts: ZodOptional<ZodObject<{
     abstain: ZodNumber;
     approve: ZodNumber;
     reject: ZodNumber;
     totalWeight: ZodNumber;
  }, $strip>>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:206

Consensus result schema for validation.

CorrelationCoefficientSchema

const CorrelationCoefficientSchema: ZodNumber;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:51

Correlation coefficient between two agents’ voting patterns. Range: -1 (perfectly anti-correlated) to +1 (perfectly correlated). 0 indicates independence.

CorrelationTrackerStatsSchema

const CorrelationTrackerStatsSchema: ZodObject<{
  averageCorrelation: ZodNumber;
  independentSubsetCount: ZodNumber;
  pairsWithSufficientData: ZodNumber;
  totalAgents: ZodNumber;
  totalObservations: ZodNumber;
  trackedPairs: ZodNumber;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:305

DEFAULT_CONSENSUS_CONFIG

const DEFAULT_CONSENSUS_CONFIG: ConsensusEngineConfig;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:344

Default configuration values.

DEFAULT_HIGHER_ORDER_CONFIG

const DEFAULT_HIGHER_ORDER_CONFIG: HigherOrderVotingConfig;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:228

DEFAULT_VOTING_PROTOCOL_CONFIG

const DEFAULT_VOTING_PROTOCOL_CONFIG: VotingProtocolConfig;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:110

DEFAULT_WEIGHTED_VOTING_CONFIG

const DEFAULT_WEIGHTED_VOTING_CONFIG: WeightedVotingConfig;

Defined in: packages/nexus-agents/src/consensus/types-weighted-voting.ts:99

FindingVoteSchema

const FindingVoteSchema: ZodObject<{
  agentId: ZodString;
  agree: ZodBoolean;
  amendedSeverity: ZodOptional<ZodEnum<{
     critical: "critical";
     major: "major";
     minor: "minor";
     suggestion: "suggestion";
  }>>;
  findingId: ZodString;
  reasoning: ZodOptional<ZodString>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:59

Finding vote during deliberation.

HigherOrderVotingConfigSchema

const HigherOrderVotingConfigSchema: ZodObject<{
  correlationMaxAgeMs: ZodDefault<ZodNumber>;
  correlationThreshold: ZodDefault<ZodNumber>;
  fallbackToSimpleVoting: ZodDefault<ZodBoolean>;
  independenceThreshold: ZodDefault<ZodNumber>;
  maxObservationsPerAgent: ZodDefault<ZodNumber>;
  maxProposals: ZodDefault<ZodNumber>;
  maxTrackedPairs: ZodDefault<ZodNumber>;
  minObservationsForCorrelation: ZodDefault<ZodNumber>;
  observationDecayFactor: ZodDefault<ZodNumber>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:210

HigherOrderVotingResultSchema

const HigherOrderVotingResultSchema: ZodObject<{
  decision: ZodEnum<{
     approve: "approve";
     no_consensus: "no_consensus";
     reject: "reject";
  }>;
  downweightedAgents: ZodArray<ZodString>;
  effectiveVoteCount: ZodNumber;
  improvementOverBaseline: ZodNumber;
  independentSubsets: ZodOptional<ZodArray<ZodObject<{
     agentIds: ZodArray<ZodString>;
     id: ZodString;
     independenceScore: ZodNumber;
     observationCount: ZodNumber;
     pairCoverage: ZodObject<{
        observed: ZodNumber;
        total: ZodNumber;
     }, $strip>;
  }, $strip>>>;
  method: ZodEnum<{
     isp: "isp";
     ow: "ow";
     simple: "simple";
  }>;
  posteriorApproval: ZodNumber;
  posteriorRejection: ZodNumber;
  reasoning: ZodString;
  usedCorrelationData: ZodBoolean;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:270

IndependentSubsetSchema

const IndependentSubsetSchema: ZodObject<{
  agentIds: ZodArray<ZodString>;
  id: ZodString;
  independenceScore: ZodNumber;
  observationCount: ZodNumber;
  pairCoverage: ZodObject<{
     observed: ZodNumber;
     total: ZodNumber;
  }, $strip>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:91

PairwiseVotingHistorySchema

const PairwiseVotingHistorySchema: ZodObject<{
  agreements: ZodNumber;
  correlation: ZodNumber;
  disagreements: ZodNumber;
  jointObservations: ZodNumber;
  lastUpdated: ZodDate;
  pairKey: ZodType<`${string}:${string}`, unknown, $ZodTypeInternals<`${string}:${string}`, unknown>>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:165

ProposalSchema

const ProposalSchema: ZodObject<{
  algorithm: ZodEnum<{
     higher_order: "higher_order";
     opinion_wise: "opinion_wise";
     proof_of_learning: "proof_of_learning";
     simple_majority: "simple_majority";
     supermajority: "supermajority";
     unanimous: "unanimous";
  }>;
  createdAt: ZodOptional<ZodISODateTime>;
  description: ZodString;
  id: ZodOptional<ZodString>;
  metadata: ZodOptional<ZodRecord<ZodString, ZodUnknown>>;
  requiredVoters: ZodOptional<ZodArray<ZodString>>;
  timeout: ZodOptional<ZodNumber>;
  title: ZodString;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:119

A proposal submitted for consensus.

ProposalStatusSchema

const ProposalStatusSchema: ZodEnum<{
  approved: "approved";
  closed: "closed";
  pending: "pending";
  rejected: "rejected";
  timeout: "timeout";
  voting: "voting";
}>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:39

Proposal status in the lifecycle.

REJECTION_CATEGORIES

const REJECTION_CATEGORIES: (
  | "YAGNI"
  | "DRY_VIOLATION"
  | "OVER_ENGINEERING"
  | "SCOPE_CREEP"
  | "SECURITY_RISK"
  | "MISALIGNED"
  | "INSUFFICIENT_EVIDENCE")[] = RejectionCategorySchema.options;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:67

All valid rejection category values, for runtime reference.

RejectionCategorySchema

const RejectionCategorySchema: ZodEnum<{
  DRY_VIOLATION: "DRY_VIOLATION";
  INSUFFICIENT_EVIDENCE: "INSUFFICIENT_EVIDENCE";
  MISALIGNED: "MISALIGNED";
  OVER_ENGINEERING: "OVER_ENGINEERING";
  SCOPE_CREEP: "SCOPE_CREEP";
  SECURITY_RISK: "SECURITY_RISK";
  YAGNI: "YAGNI";
}>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:53

Structured rejection feedback categories (Issue #1213). Enables reject→refine→re-vote feedback loops by classifying rejection reasons.

VoteSchema

const VoteSchema: ZodObject<{
  conditions: ZodOptional<ZodArray<ZodString>>;
  confidence: ZodNumber;
  decision: ZodEnum<{
     abstain: "abstain";
     approve: "approve";
     reject: "reject";
  }>;
  findings: ZodOptional<ZodArray<ZodObject<{
     claim: ZodString;
     gate: ZodObject<{
        named_assertion: ZodDefault<ZodString>;
        reread_cited_line: ZodDefault<ZodEnum<{
           failed: ...;
           passed: ...;
           skipped: ...;
        }>>;
        ruled_out_language_non_issue: ZodDefault<ZodEnum<{
           failed: ...;
           passed: ...;
           skipped: ...;
        }>>;
        traced_call_path: ZodDefault<ZodEnum<{
           failed: ...;
           passed: ...;
           skipped: ...;
        }>>;
     }, $strip>;
     location: ZodString;
     severity: ZodDefault<ZodEnum<{
        critical: "critical";
        high: "high";
        low: "low";
        medium: "medium";
     }>>;
     summary: ZodString;
  }, $strip>>>;
  reasoning: ZodString;
  rejectionCategories: ZodOptional<ZodArray<ZodEnum<{
     DRY_VIOLATION: "DRY_VIOLATION";
     INSUFFICIENT_EVIDENCE: "INSUFFICIENT_EVIDENCE";
     MISALIGNED: "MISALIGNED";
     OVER_ENGINEERING: "OVER_ENGINEERING";
     SCOPE_CREEP: "SCOPE_CREEP";
     SECURITY_RISK: "SECURITY_RISK";
     YAGNI: "YAGNI";
  }>>>;
  selectedOption: ZodOptional<ZodString>;
  timestamp: ZodOptional<ZodISODateTime>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-core.ts:91

A vote cast by an agent.

VOTING_THRESHOLDS

const VOTING_THRESHOLDS: Record<ConsensusAlgorithm, number>;

Defined in: packages/nexus-agents/src/consensus/decision/thresholds.ts:29

Voting thresholds for each algorithm.

VotingObservationSchema

const VotingObservationSchema: ZodObject<{
  agentId: ZodString;
  alignedWithOutcome: ZodBoolean;
  confidence: ZodNumber;
  decision: ZodEnum<{
     abstain: "abstain";
     approve: "approve";
     reject: "reject";
  }>;
  modelKey: ZodOptional<ZodString>;
  observedModel: ZodOptional<ZodString>;
  proposalId: ZodString;
  timestamp: ZodDate;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:128

VotingProtocolConfigSchema

const VotingProtocolConfigSchema: ZodObject<{
  agreementThreshold: ZodDefault<ZodNumber>;
  committeeSize: ZodDefault<ZodNumber>;
  enableAntiSycophancy: ZodDefault<ZodBoolean>;
  maxRounds: ZodDefault<ZodNumber>;
  roundTimeoutMs: ZodDefault<ZodNumber>;
  sycophancyThreshold: ZodDefault<ZodNumber>;
}, $strip>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:101

VotingRoundPhaseSchema

const VotingRoundPhaseSchema: ZodEnum<{
  analysis: "analysis";
  consensus: "consensus";
  deliberation: "deliberation";
}>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:26

Voting round phases.

analysis: Independent analysis (Round 1)
deliberation: Share findings and discuss (Round 2)
consensus: Final vote on recommendations (Round 3)

VotingRoundStatusSchema

const VotingRoundStatusSchema: ZodEnum<{
  aborted: "aborted";
  awaiting_votes: "awaiting_votes";
  completed: "completed";
  in_progress: "in_progress";
  pending: "pending";
}>;

Defined in: packages/nexus-agents/src/consensus/types-voting-protocol.ts:32

Voting round status.

Functions

buildFinalResult()

function buildFinalResult(
   state, 
   proposalId, 
   outcome, 
   config
): ConsensusResult;

Defined in: packages/nexus-agents/src/consensus/result-builder.ts:54

Build a final result for a closed proposal.

Parameters

state

ProposalState

proposalId

string

outcome

config

ConsensusEngineConfig

Returns

ConsensusResult

buildPendingResult()

function buildPendingResult(
   state, 
   proposalId, 
   outcome, 
   config
): ConsensusResult;

Defined in: packages/nexus-agents/src/consensus/result-builder.ts:28

Build a pending result for an active proposal.

Parameters

state

ProposalState

proposalId

string

outcome

config

ConsensusEngineConfig

Returns

ConsensusResult

buildTimeoutResult()

function buildTimeoutResult(
   state, 
   proposalId, 
   outcome, 
   config
): ConsensusResult;

Defined in: packages/nexus-agents/src/consensus/result-builder.ts:83

Build a timeout result for an expired proposal.

Parameters

state

ProposalState

proposalId

string

outcome

CollectRealVotesOptions

config

ConsensusEngineConfig

Returns

ConsensusResult

calculateVoteWeight()

function calculateVoteWeight(performance): number;

Defined in: packages/nexus-agents/src/consensus/strategies.ts:339

Calculate vote weight for an agent based on their performance history.

Weight ranges from 0.5 (never correct) to 1.0 (perfect track record). An agent with NO history returns 1.0, not 0.5 — the doc used to say “0.5 (no history)”, which the code has never done (#5117). The distinction matters: a 1.0 default is indistinguishable from a perfect record by value alone, which is why callers must not infer “was this measured?” from the number. deriveWeightBasis answers that from provenance instead.

Parameters

performance

AgentPerformance | undefined

Returns

number

collectRealVotes()

function collectRealVotes(options): Promise<readonly AgentVoteResult[]>;

Defined in: packages/nexus-agents/src/cli/voter-agents.ts:580

Parameters

options

Returns

Promise<readonly AgentVoteResult[]>

createAgentPairKey()

function createAgentPairKey(agentA, agentB): `${string}:${string}`;

Defined in: packages/nexus-agents/src/consensus/higher-order-types.ts:31

Creates a canonical agent pair key for correlation lookup. Orders agents lexicographically to ensure consistent keys.

Parameters

agentA

string

agentB

string

Returns

`${string}:${string}`

createConsensusEngine()

function createConsensusEngine(config?, logger?): ConsensusEngine;

Defined in: packages/nexus-agents/src/consensus/engine.ts:721

Create a consensus engine with the given configuration.

Parameters

config?

Partial<ConsensusEngineConfig>

logger?

ILogger

Returns

ConsensusEngine

Example

const engine = createConsensusEngine({
  defaultTimeout: 60000,
  maxActiveProposals: 10,
});

createCorrelationTracker()

function createCorrelationTracker(config?): ICorrelationTracker;

Defined in: packages/nexus-agents/src/consensus/correlation-tracker.ts:385

Creates a new correlation tracker instance.

Parameters

config?

Partial<HigherOrderVotingConfig>

Returns