Releases: mehdi-arfaoui/Stronghold
Stronghold v1.2.0 - Full-Chain Recovery Coverage
What's new
Full-Chain Recovery Coverage
Stronghold now traces the complete recovery path for each service
and evaluates every step individually.
stronghold explain databaseRecovery Chain — database (0/5 proven · weighted 0%)
Step 1 ✗ prod-db-primary restore BLOCKED (no backup plan)
Step 2 ✗ prod-db-replica failover BLOCKED (no replica exists)
Step 3 ~ payment-cache warmup OBSERVED (never tested)
Step 4 ~ payment-api redeploy OBSERVED (never tested)
Step 5 ~ healthcheck validation OBSERVED (never tested)
Each step is classified as proven (tested), observed (config seen
but never tested), blocked (failing rules), or unknown. Steps are
weighted by role — an unproven datastore is more critical than an
unproven compute node.
Install
npx @stronghold-dr/cli@1.2.0 demo
npx @stronghold-dr/cli@1.2.0 explain database
npx @stronghold-dr/cli@1.2.0 graphStronghold v1.1.0 - Reality Gap Engine
What's new
Reality Gap
Stronghold now measures the distance between what config checks report
and what is actually proven recoverable.
Reality Gap 56 pts claimed 56% protected → 0% proven recoverable
stronghold explain
Full reasoning chains show why a service has its score — step by step,
deterministic, verifiable, no LLM.
Graph Insights
Four graph-based intelligence types that surface risks invisible to
manual review: cascade failure detection, silent dependency drift,
risk acceptance invalidation, and recovery path erosion.
Interactive Graph
stronghold graph generates a standalone HTML with dependency
visualization, Reality Gap bar, scenario mode, and reasoning sidebar.
Install
npx @stronghold-dr/cli@1.1.0 demo
npx @stronghold-dr/cli@1.1.0 explain database
npx @stronghold-dr/cli@1.1.0 graphStronghold v1.0.0
Stronghold v1.0.0
The first stable release of Stronghold -- open-source disaster recovery intelligence for AWS.
Install
npx @stronghold-dr/cli demo
npx @stronghold-dr/cli scan --region eu-west-1What's in v1.0.0
Service-centric DR intelligence. Stronghold maps AWS resources into services and reasons about recoverability at the service level, not just the resource level.
Evidence-backed posture. Five evidence maturity levels (observed, inferred, declared, tested, expired) with confidence-weighted scoring. Know the difference between "config exists" and "recovery was proven."
Scenario coverage analysis. Built-in AZ failure, region failure, SPOF failure, and data corruption scenarios. Coverage verdicts tell you which services survive which disruptions.
Living runbooks. Generated DR plans and executable runbooks validated against current infrastructure. Stale recovery references are flagged before an incident exposes them.
Posture memory. Scan history, finding lifecycle tracking, DR debt, and trend analysis. See whether your DR posture is improving, stable, or degrading.
Lightweight governance. Declared ownership, risk acceptance with mandatory expiration, and custom policy enforcement.
Infrastructure
- Read-only AWS discovery across 16 services with bounded concurrency, retries, and per-scanner timeouts
- Multi-account support with
--profile,--role-arn, and named account configuration - AES-256-GCM encryption, redaction engine, and always-on audit trail
DR Engine
- 39 validation rules across 6 DR categories
- Weighted scoring with severity ceiling, role-aware weighting, and evidence maturity
- DRP-as-Code generation with topological recovery ordering
- Executable runbooks with real AWS CLI commands and honest RTO/RPO
Operations
- Drift detection between scans with DRP impact analysis
- CI-native integration with
--format json,--output json,--fail-threshold, and GitHub Actions annotations - Demo mode with 3 scenarios (startup, enterprise, minimal)
- Self-hosted deployment with Docker Compose (Express + React + PostgreSQL)
Documentation
- Getting Started
- Architecture
- Service Model
- Evidence Model
- Scenario Coverage
- Governance
- Posture History
- Security Model
- Scoring Methodology
Technical Foundation
- TypeScript strict mode, zero
any,noUncheckedIndexedAccess: true - Monorepo:
@stronghold-dr/core(pure business logic) +@stronghold-dr/cli+ server + web - 654 passing automated tests, 81.62% core line coverage
- AGPL-3.0