Analyzing Sbom For Supply Chain VulnerabilitiesSAFE

817 structured cybersecurity skills for AI agents · Mapped to 6 frameworks: MITRE ATT&CK, NIST CSF 2.0, MITRE ATLAS, D3FEND, NIST AI RMF & MITRE F3 (Fight Fraud) · agentskills.io standard · Works with Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI & 20+ platforms · 29 security domains ·

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: analyzing-sbom-for-supply-chain-vulnerabilities
description: 'Parses Software Bill of Materials (SBOM) in CycloneDX and SPDX JSON
  formats to identify supply chain vulnerabilities by correlating components against
  the NVD CVE database via the NVD 2.0 API. Builds dependency graphs, calculates risk
  scores, identifies transitive vulnerability paths, and generates compliance reports.
  Activates for requests involving SBOM analysis, software composition analysis, supply
  chain security assessment, dependency vulnerability scanning, CycloneDX/SPDX parsing,
  or CVE correlation.

  '
domain: cybersecurity
subdomain: supply-chain-security
tags:
- SBOM
- CycloneDX
- SPDX
- NVD
- CVE
- supply-chain
- dependency-analysis
- syft
- grype
version: 1.0.0
author: mukul975
license: Apache-2.0
atlas_techniques:
- AML.T0010
nist_ai_rmf:
- GOVERN-5.2
- MAP-1.6
- MANAGE-2.2
- GOVERN-1.1
- GOVERN-4.2
nist_csf:
- GV.SC-01
- GV.SC-03
- GV.SC-06
- GV.SC-07
mitre_attack:
- T1195.001
- T1195.002
- T1554
- T1190
---

# Analyzing SBOM for Supply Chain Vulnerabilities

## When to Use

- A new regulatory requirement (EO 14028, EU CRA) mandates SBOM analysis for software deliveries
- Security team needs to assess third-party risk by scanning vendor-provided SBOMs
- CI/CD pipeline requires automated vulnerability checks against generated SBOMs
- Incident response needs to determine if a newly disclosed CVE affects deployed software
- Procurement team requires supply chain risk assessment for a software acquisition

**Do not use** for runtime vulnerability scanning of live systems; use container scanning tools (Trivy, Grype CLI) or host-based vulnerability scanners (Nessus, Qualys) instead.

## Prerequisites

- SBOM file in CycloneDX JSON (v1.4+) or SPDX JSON (v2.3+) format
- Python 3.9+ with requests, networkx, and packaging libraries installed
- NVD API key (free, from https://nvd.nist.gov/developers/request-an-api-key) for higher rate limits
- Network access to NVD API (https://services.nvd.nist.gov/rest/json/cves/2.0)
- Optionally: syft for SBOM generation, grype for cross-validation

## Workflow

### Step 1: Generate SBOM (if not provided)

Use syft to create an SBOM from a container image or project directory:

```bash
# Generate CycloneDX JSON from a container image
syft alpine:latest -o cyclonedx-json > sbom-cyclonedx.json

# Generate SPDX JSON from a project directory
syft dir:/path/to/project -o spdx-json > sbom-spdx.json

# Generate from a running container
syft docker:my-app-container -o cyclonedx-json > sbom.json
```

Syft supports over 30 package ecosystems including npm, PyPI, Maven, Go modules, apt, apk, and RPM. The generated SBOM includes package names, versions, licenses, CPE identifiers, and PURL (Package URL) references.

### Step 2: Parse SBOM and Extract Components

Parse the SBOM to extract all software components with their identifiers:

**CycloneDX JSON Structure:**
```json
{
  "bomFormat": "CycloneDX",
  "specVersion": "1.5",
  "components": [
    {
      "type": "library",
      "name": "lodash",
      "version": "4.17.20",
      "purl": "pkg:npm/[email protected]",
      "cpe": "cpe:2.3:a:lodash:lodash:4.17.20:*:*:*:*:*:*:*",
      "licenses": [{"license": {"id": "MIT"}}]
    }
  ],
  "dependencies": [
    {"ref": "pkg:npm/[email protected]", "dependsOn": ["pkg:npm/[email protected]"]}
  ]
}
```

**SPDX JSON Structure:**
```json
{
  "spdxVersion": "SPDX-2.3",
  "packages": [
    {
      "name": "lodash",
      "versionInfo": "4.17.20",
      "externalRefs": [
        {"referenceType": "purl", "referenceLocator": "pkg:npm/[email protected]"},
        {"referenceType": "cpe23Type", "referenceLocator": "cpe:2.3:a:lodash:lodash:4.17.20:*:*:*:*:*:*:*"}
      ],
      "licenseConcluded": "MIT"
    }
  ],
  "relationships": [
    {"spdxElementId": "SPDXRef-express", "relatedSpdxElement": "SPDXRef-lodash",
     "relationshipType": "DEPENDS_ON"}
  ]
}
```

### Step 3: Correlate Components with NVD CVE Database

Query the NVD 2.0 API to find known vulnerabilities for each component:

```python
import requests

NVD_API = "https://services.nvd.nist.gov/rest/json/cves/2.0"

def search_cves_by_cpe(cpe_name, api_key=None):
    params = {"cpeName": cpe_name, "resultsPerPage": 50}
    headers = {"apiKey": api_key} if api_key else {}
    resp = requests.get(NVD_API, params=params, headers=headers, timeout=30)
    resp.raise_for_status()
    return resp.json().get("vulnerabilities", [])

def search_cves_by_keyword(keyword, version=None, api_key=None):
    params = {"keywordSearch": keyword, "resultsPerPage": 50}
    headers = {"apiKey": api_key} if api_key else {}
    resp = requests.get(NVD_API, params=params, headers=headers, timeout=30)
    resp.raise_for_status()
    return resp.json().get("vulnerabilities", [])
```

The NVD API supports searching by CPE name (most precise), keyword, CVE ID, and date ranges. Rate limits: 5 requests/30 seconds without API key, 50 requests/30 seconds with key.

### Step 4: Build Dependency Graph and Identify Transitive Risks

Construct a directed graph of dependencies to trace vulnerability propagation:

```python
import networkx as nx

def build_dependency_graph(sbom):
    G = nx.DiGraph()
    # Add nodes for each component
    for comp in sbom["components"]:
        G.add_node(comp["purl"], name=comp["name"], version=comp["version"])
    # Add edges from dependency relationships
    for dep in sbom.get("dependencies", []):
        for child in dep.get("dependsOn", []):
            G.add_edge(dep["ref"], child)
    return G
```

Transitive dependency analysis identifies components that are not directly included but are pulled in through dependency chains. A vulnerability in a deeply nested transitive dependency (e.g., 4 levels deep) still represents risk but may be harder to remediate.

Key graph metrics for risk assessment:
- **In-degree**: How many components depend on this one (high in-degree = high blast radius)
- **Shortest path to root**: Distance from application

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

declared (3 observation(s))

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (2)

curl -sSfL https://raw.githubusercontent.com/anchore/syft/main/install.sh | sh -s -- -b /usr/local/bin

curl -sSfL https://raw.githubusercontent.com/anchore/grype/main/install.sh | sh -s -- -b /usr/local/bin

Gates applied: no_behavioural_pass.