Version Matrix

July 25, 2026 · View on GitHub

This file is the canonical reference for AdversaryGraph release history and feature gates.

Current Release

FieldValue
Versionv6.5.0
Release date2026-07-25
ThemeGoverned intelligence, hunting, exposure assessment, and SOC operations
StatusSource release prepared; immutable tag and artifacts require the successful tag workflow

The source release is v6.5.0. The latest previously published immutable tag is v6.0.0 until the v6.5.0 tag workflow completes. Historical v6.0 screenshots and artifact limitations remain scoped to that tag.

v6.5 Capability Promotion

v6.5.0 promotes the complete development line after v6.0.0: governed Threat Hunting and Query Library workflows, multi-provider AI, unified hybrid RAG, the Navigator intelligence assistant, local stdio MCP, saved-asset intelligence, inventory-bound exposure assessment, persistent SOC access groups, module-level authorization across 31 workspaces, complete API contracts, and the hardened seven-image release path. They are documented in Unified Intelligence RAG and MCP, the v6.5.0 release notes, and the changelog. They are not retroactively attributed to the immutable v6.0.0 tag. The earlier v6.1.0 source milestone was not published as a stable tag and is superseded by v6.5.0.

Release History

VersionThemeKey additions
v6.5.0Governed Intelligence, Hunting, Exposure Assessment, and SOC OperationsThreat Hunting and Query Library workflows, unified RAG/MCP, saved-asset intelligence, inventory-bound passive/Nmap/web assessment, persistent SOC groups, module-level API/UI authorization across 31 workspaces, complete API contracts, and post-v6 platform hardening
v6.0.0Operational Evidence and Production ReadinessReproducible release gate, corrected v5 history, tagged screenshot evidence, local case studies, deployment go/no-go criteria, version-derived UI metadata, and reviewer handoff material
v5.9.1JA3/JA4+ Network Fingerprint IOC WorkflowsJA3/JA3S/JA4/JA4S/JA4H/JA4L/JA4LS/JA4X/JA4SSH/JA4T IOC types, report-text extraction, normalized import tagging, IOC Library filtering, IOC Detail context, IOC node detail support, and IOC Investigation pivots
v5.9.0EMB3D and Threat Radar Asset WorkflowsEMB3D API/service/UI/documentation, unified product/component/dependency/asset modeling, full asset-inventory import templates, product-security sample datasets, and Threat Radar asset review pages
v5.8.0Threat Radar Product-Security CTIThreat Radar module, scored CVE/KEV/PoC/zero-day/supplier/package/hardware/customer/internal telemetry signals, product/component/dependency exposure mappings, case graph, sanitized legal-sensitive evidence handling, PSIRT/Hunt/IR/Detection queues, watchlists, and generated reports
v5.7.0Research Collection and Linked Report ReviewReports / Research collection page, linked report review with inline entity links, source-text preservation for AI analysis, store-only research upload, Parse with AI upload workflow, and research analysis guide
v5.6.0Statistics Tag AnalyticsExpanded Statistics module with IOC/CVE/TTP/actor/report/sector/global tag widgets for risk, confidence, region, sector, type, source, telemetry, TLP, attack vector, malware family, and relationship-confidence analysis
v5.5.0Enterprise Access ControlsExpanded RBAC roles, per-user permissions, password policy settings, MFA workflow support, trusted proxy SSO metadata, session inventory and revocation, authentication audit history, Admin Panel updates, and deployment configuration coverage
v5.4.0Observability and Validation EvidenceAuthenticated Observability dashboard, request metrics, recent traces, redacted API log tail, Prometheus-compatible metrics endpoint, backend SAST CI coverage, security scan helper, and screenshot-backed validation examples
v5.3.0Authentication and User OperationsLocal /auth-guide page reachable before sign-in, login-page guide link, native auth bootstrap guidance, role model documentation, password reset/session behavior notes, and production/security docs for native auth plus optional identity-aware reverse proxy
v5.2.0QA Hardening and Release ValidationReproducible backend test environment defaults, frontend DOMPurify override for Monaco transitive audit cleanup, local lint/test/audit/build validation, and v5.2 release metadata
v5.1.0Telemetry Fidelity, Raw STIX, and CVE Library CorrelationSource-correct telemetry policy for Attack Simulation, raw STIX object/relationship preservation, CVE Library with NVD/CISA KEV sync, CVSS score fields, and strict APT-TTP-IOC-CVE links, AI assistant prompt guardrails, updated architecture documentation, CI-validated release metadata
v5.0.0Attack Simulation and SIEM ValidationTTP-first simulation matrix, real lab-target attack flows, AI kill-chain telemetry generation, SIEM forwarding with authentication, Scenario Library, attack-chain graph view
v4.1.0Detection CoverageDetection coverage states per technique, Sigma/KQL/SPL/EQL skeleton export, telemetry source tracking, coverage summaries by tactic and platform
v4.0.0Detection Engineering WorkflowDetection backlog export, detection coverage tracking, production-readiness hardening
v3.2.0Evidence BindingSource paragraph/span references, evidence snippets beside ATT&CK mappings, evidence-backed export
v3.1.0Analyst Review WorkflowReview states (suggested/accepted/rejected/needs-evidence), analyst notes, confidence filtering
v3.0.0Malware Analysis ModuleYARA scanning, string extraction, PE header parsing, IOC extraction, AI-assisted analysis
v2.xReport ProcessingMulti-format ingestion, AI TTP extraction, ATT&CK mapping, Navigator export, JSONB storage
v0.2.0–v1.xFoundationInitial FastAPI backend, React frontend, PostgreSQL, Redis, Celery, Docker Compose

For complete per-version changelogs see CHANGELOG.md. For a consolidated account of every v5 release, see the v5 overview. For the current release narrative, see v6.5.0 release notes.

Feature Gate Legend

LabelMeaning
ImplementedShipped and available in the current release
Implemented (partial)Core logic shipped; some UI controls or edge cases remain pending
PlannedOn the roadmap but not yet started
GatedAvailable only in specific deployment configurations
AI-generatedOutput is produced by an LLM and requires analyst review before use
SyntheticTelemetry or data is generated for testing purposes, not from a real attack
Not claimedFunctionality that is sometimes assumed but is explicitly not implemented

See ROADMAP.md for upcoming work.