Skip to main content

@mellens/rave

v2026.04.02.1

RAVE (Reliability & Validation Engineering) — confidence scoring with exponential decay, claim lifecycle management, evidence gathering from GitHub and Prometheus, falsification detection, and readiness gating for software reliability claims

Labels

ravereliabilityconfidenceclaimsevidencefalsificationreadiness

Contents

Quality score

How well-documented and verifiable this extension is.

Not yet scored.

A score will be generated the next time this extension is published. The owner can also trigger scoring manually.

Install

$ swamp extension pull @mellens/rave

@mellens/rave/ci-evidencev2026.03.21.1rave_ci_evidence.ts

Global Arguments

ArgumentTypeDescription
evidenceIdstring
repostring
workflowNamestring
branchstring
gatherFetch the latest GitHub Actions workflow run and record the outcome
ArgumentTypeDescription
githubTokenstring

Resources

result(30d)— Latest CI run result for this evidence entity
@mellens/rave/claimv2026.03.22.1rave_claim.ts

Global Arguments

ArgumentTypeDescription
claimIdstring
createWrite a new claim YAML to rave/claims/ and record its initial state
ArgumentTypeDescription
statementstring
ownerstring
team?string
contact?string
categorystring
scopeTypestring
scopeTargetstring
assumptionsarray
falsificationSignalsarray
getRead the current claim YAML and record a state snapshot
activateSet claim status to 'active' — use once evidence is wired up
retireSet claim status to 'retired' — freezes the confidence score
ArgumentTypeDescription
reason?string
contradictSet claim status to 'contradicted' — collapses confidence to 0
ArgumentTypeDescription
reasonstring
annotateAdd a free-text annotation to the claim YAML
ArgumentTypeDescription
textstring
authorstring

Resources

state(infinite)— Current claim state snapshot (mirrors key fields from the YAML)
@mellens/rave/confidence-enginev2026.03.21.1rave_confidence_engine.ts

Global Arguments

ArgumentTypeDescription
claimIdstring
computeApply the RAVE decay formula to produce an updated confidence score for this claim
ArgumentTypeDescription
currentScorenumber
lastValidatedstring
currentStatusstring
evidencearray
revalidateReset the confidence anchor when a claim owner formally re-attests
ArgumentTypeDescription
newScorenumber
revalidatedBystring

Resources

confidence(infinite)— Computed confidence state for this claim
@mellens/rave/falsifier-enginev2026.03.21.1rave_falsifier_engine.ts

Global Arguments

ArgumentTypeDescription
falsifierIdstring
conditionTypeenum
conditionstring
evaluateEvaluate falsifier conditions against evidence snapshots and record triggered/not-triggered
ArgumentTypeDescription
evidencearray

Resources

evaluation(30d)— Latest falsifier evaluation result
@mellens/rave/github-api-evidencev2026.03.22.1rave_github_api_evidence.ts

Global Arguments

ArgumentTypeDescription
evidenceIdstring
repostring
endpointstring
successField?string
gatherCall a GitHub REST API endpoint and record the outcome
ArgumentTypeDescription
githubTokenstring

Resources

result(1d)— Latest GitHub REST API evidence result
@mellens/rave/prometheus-evidencev2026.03.21.1rave_prometheus_evidence.ts

Global Arguments

ArgumentTypeDescription
evidenceIdstring
baseUrlstring
querystring
threshold?number
operator?enum
unit?string
gatherExecute a PromQL instant query and record the outcome against an optional threshold
ArgumentTypeDescription
prometheusTokenstring

Resources

result(7d)— Latest Prometheus query result
@mellens/rave/readiness-reporterv2026.03.23.1rave_readiness_reporter.ts
evaluateEvaluate readiness across all active claims by comparing confidence scores to threshold. Draft and retired claims are excluded; contradicted claims always fail.
ArgumentTypeDescription
branchProtectionIdstring
branchProtectionStatusstring
ciGreenIdstring
ciGreenStatusstring
swampModelsIdstring
swampModelsStatusstring
swampWorkflowsIdstring
swampWorkflowsStatusstring
extensionsCompileIdstring
extensionsCompileStatusstring

Resources

report(7d)— Latest readiness evaluation report