Current version: v2.3.0 (May 2026)
- 17 curated AWS security checks (IAM, S3, EC2, VPC, RDS, EIP)
- Rich CLI with progress bar and colored output
- JSON and HTML report output
- Health score (0-100) based on finding severity
- Docker support
- Every finding includes copy-paste remediation (AWS CLI + Terraform HCL + docs link)
- CIS AWS Foundations Benchmark references on applicable checks
--export-fixesgenerates a commented shell script for safe bulk remediation- Effort estimation per finding (LOW / MEDIUM / HIGH)
- moto-based test suite with 80%+ coverage
- CloudTrail, GuardDuty, AWS Config, KMS, CloudWatch checks
- Total: 27 checks
- Lambda, ECS, SSM, Secrets Manager checks
- Total: 42 checks
- SARIF output, config file, suppressions, env vars, OIDC, markdown output
- Total: 45 checks
- SHA-pinned GitHub Actions, CI gate for release, non-root Docker, GHCR
- Total: 45 checks
- SARIF spec compliance, S3 encryption pivot, accessibility, ruff rules
- Total: 45 checks
cloud-audit diffcommand for tracking drift- CI/CD examples (daily-scan-with-diff, post-deploy-scan)
- Breach cost estimation (IBM, Verizon data), MCP Server
- 278 tests
- CIS AWS v3.0: 62 controls, 55 automated
- MkDocs documentation site, 303 tests
- 43 Trust Services Criteria, 335 tests
- BSI C5:2020, ISO 27001:2022, HIPAA Security Rule, NIS2 Directive
- 6 compliance frameworks total, 412 tests
- IAM Privilege Escalation Detection -- 25 methods across 6 categories (self-mutation, credential access, PassRole+service, Lambda code mod, trust policy abuse, permission boundary bypass). First maintained open-source replacement for PMapper.
- What-If Remediation Simulator --
cloud-audit simulate --fix aws-vpc-002shows impact on score, chains, risk before changing anything in AWS - Root Cause Grouping -- "fix 4 things, break 22 chains" prioritization. Groups findings by root cause and ranks by chain-breaking impact.
- Security Posture Trend --
cloud-audit trendtracks health score, chains, and risk over time. History auto-saved after each scan. - AI-SPM (Bedrock + SageMaker) -- 5 checks, 3 attack chains (model theft, LLMjacking, data poisoning). First open-source AI-SPM scanner.
- Quick Wins -- CLI section showing LOW-effort fixes that break CRITICAL chains
- 6 new attack chain rules (AC-34 through AC-39)
- Compliance Beta labels (CIS + SOC2 stable, 4 others beta)
- Remediation placeholders replaced with real values
- Terraform snippets completed with missing dependencies
- Cached get_account_id() (1 STS call instead of 10+)
- 94 checks, 23 services, 31 chains, 496 tests
- Blast Radius CLI (MVP) --
cloud-audit blast-radius --resource <id>walks outward from a single AWS resource (EC2, IAM Role/User, Lambda, S3 bucket, secret) and reports what an attacker could reach if that resource were compromised. Tree, JSON (BlastRadiusGraph v1.0 schema), Mermaid, and Markdown output. Pure in-memory, no AWS calls at blast-radius time. Built on the existing IAM trust graph + escalation catalog. 26 new tests, 812 total. Includes a full pre-release security pass (SEC-001 through SEC-009 plus F-S2-01 through F-S2-06 hardening — URL scheme allowlisting in HTML reports, symlink-safe writes, markdown injection escape, cycle dedup in BFS).
1. Security Graph + Effective Exposure Score In-memory graph (networkx) modeling all resource relationships: VPC routing, subnets, security groups, EC2 instances, IAM roles, policies, S3 buckets, RDS instances. BFS/DFS from internet nodes to high-value targets. Per-resource "effective exposure score" combining network reachability + identity privilege + data sensitivity. Output: "3 paths from internet to production database."
2. Blast Radius Analysis (full v3.0.0 expansion)
MVP shipped in v2.3.0. Remaining for v3.0.0: network reachability (VPC peering, TGW, on-prem CIDR), cross-account propagation, permission boundary / SCP semantic evaluation, data classification (PII / PCI / PHI tagging on buckets), expanded storytelling templates, HTML report integration, MCP server compute_blast_radius tool.
3. NHI (Non-Human Identity) Audit + Data Perimeter Scanner Full inventory of all non-human identities: IAM users with access keys, IAM roles, OIDC providers, Lambda/ECS/CodeBuild execution roles. Aging, rotation, privilege scoring, trust chain mapping. Data perimeter: RCPs, VPC endpoint policies, aws:SourceOrgID conditions, snapshot sharing.
- Multi-account scanning (AWS Organizations)
- SCP + permission boundary evaluation in IAM escalation
- Terraform drift detection (live state vs IaC)
- Cross-account attack path detection
- Terraform remediation.tf generation (complete, apply-ready)
- Python plugin API for custom checks
- Evidence package generation for auditors (ZIP with API dumps per control)