Agent Bom

agent-bom — AI Agent Infrastructure Security Scanner

Discovers MCP clients and servers across 20+ AI tools, scans for CVEs, maps blast radius, runs cloud CIS benchmarks, checks OWASP/NIST/MITRE compliance, generates SBOMs, and assesses AI infrastructure against AISVS v1.0 and MAESTRO framework layers.

Install

pipx install agent-bom
agent-bom scan              # auto-discover + scan
agent-bom check langchain   # check a specific package
agent-bom where             # show all discovery paths

As an MCP Server

{
  "mcpServers": {
    "agent-bom": {
      "command": "uvx",
      "args": ["agent-bom", "mcp"]
    }
  }
}

Tools (31)

Vulnerability Scanning

Compliance & Policy

Registry & Trust

Runtime & Analytics

Specialized Scans

Resources

Example Workflows

# Check a package before installing
check(package="@modelcontextprotocol/server-filesystem", ecosystem="npm")

# Map blast radius of a CVE
blast_radius(cve_id="CVE-2024-21538")

# Full scan
scan()

# Run CIS benchmark
cis_benchmark(provider="aws")

# Run AISVS v1.0 compliance
aisvs_benchmark()

# Scan vector databases for auth misconfigurations
vector_db_scan()

# Discover GPU containers, K8s GPU nodes, and unauthenticated DCGM endpoints
gpu_infra_scan()

# Assess trust of a skill file
skill_trust(skill_content="<paste SKILL.md content>")

Guardrails

Always do:

• Show CVEs even when NVD analysis is pending or severity is unknown — a CVE ID with no details is still a real finding. Report what is known; mark severity as unknown explicitly.
• Confirm with the user before scanning cloud environments (cis_benchmark) — these make live API calls to AWS/Azure/GCP using the user's credentials.
• Treat UNKNOWN severity as unresolved, not benign — it means data is not yet available, not that the issue is minor.

Never do:

• Do not modify any files, install packages, or change system configuration. This skill is read-only.
• Do not transmit env var values, credentials, or file contents to any external service. Only package names and CVE IDs leave the machine.
• Do not invoke scan() autonomously on sensitive environments without user confirmation. The autonomous_invocation policy is restricted.

Stop and ask the user when:

• The user requests a cloud CIS benchmark and no cloud credentials are configured.
• A scan finds CRITICAL CVEs — present findings and ask whether to generate a remediation plan.
• The user asks to scan a path outside their home directory.

Supported Frameworks

• OWASP LLM Top 10 (2025) — prompt injection, supply chain, data leakage
• OWASP Agentic Top 10 — tool poisoning, rug pulls, credential theft
• OWASP AISVS v1.0 — AI Security Verification Standard (9 checks)
• MITRE ATLAS — adversarial ML threat framework
• MITRE ATT&CK Enterprise — cloud/infra T-code mapping on CIS failures
• MAESTRO — KC1–KC6 layer tagging on all findings
• EU AI Act — risk classification, transparency, SBOM requirements
• NIST AI RMF — govern, map, measure, manage lifecycle
• CIS Foundations — AWS, Azure v3.0, GCP v3.0, Snowflake benchmarks

Privacy & Data Handling

This skill installs agent-bom from PyPI. The redaction behavior described here is implemented in the installed package — verify before running with sensitive data:

# 1. Verify package integrity (Sigstore)
agent-bom verify agent-bom

# 2. Review the redaction code directly
# security.py L159: sanitize_env_vars() — replaces env values with ***REDACTED***
# https://github.com/msaad00/agent-bom/blob/main/src/agent_bom/security.py#L159

# 3. Review config parsing
# https://github.com/msaad00/agent-bom/blob/main/src/agent_bom/discovery/__init__.py

Discovery reads local MCP client config files. Only server names, commands, args, and URLs are extracted. Env var values are replaced with ***REDACTED*** by sanitize_env_vars() in the installed code. Only public package names and CVE IDs are sent to vulnerability databases. Cloud CIS checks use locally configured credentials and call only the cloud provider's own APIs.

Verification

• Source: github.com/msaad00/agent-bom (Apache-2.0)
• Sigstore signed: agent-bom verify agent-bom@0.70.6
• 5,987+ tests with CodeQL + OpenSSF Scorecard
• No telemetry: Zero tracking, zero analytics