agent-bom — Security Platform for Agentic Infrastructure
Scan your MCP servers, AI agents, and their packages: CVEs and supply-chain
risk (malicious/typosquat) in the packages, plus credential exposure,
identity/auth/access posture, and MCP config risk on the servers and agents.
Maps blast radius from a finding to the credentials and tools it can reach.
When to Use
- Before installing a new MCP server or skill
- To audit your Cortex Code MCP configuration for vulnerabilities
- To check if a specific package has known CVEs
- To map the blast radius of a vulnerability across your agent setup
- To generate an SBOM for compliance (CycloneDX, SPDX)
- To verify package integrity via Sigstore provenance
What This Skill Provides
- CVE scanning with enrichment from OSV, NVD, EPSS, CISA KEV
- MCP server discovery across real AI developer tools, including Cortex Code
- Blast radius mapping — which agents, credentials, and tools are exposed
- Compliance mapping across OWASP, NIST, MITRE, EU AI Act, and related frameworks
- Runtime proxy for MCP traffic interception and policy enforcement
- SKILL.md trust assessment with 17 behavioral risk patterns
Install
# Install via pip or pipx
pipx install agent-bom
# Or run directly with uvx (no install needed)
uvx agent-bom scan
Instructions
Scan the current MCP setup
agent-bom scan
Auto-discovers all configured MCP clients (Cortex Code, Claude Desktop, Cursor, etc.),
resolves their server packages, and scans for CVEs.
Check a package before installing
agent-bom check <package>@<version> --ecosystem <npm|pypi|go|maven|cargo|nuget|gem>
Map blast radius of a CVE
agent-bom scan --enrich
# Then review the blast radius section in the output
Generate SBOM
agent-bom scan -f cyclonedx -o sbom.json
agent-bom scan -f spdx -o sbom.spdx.json
Scan with GPU infrastructure detection
agent-bom scan --gpu-scan
Run as MCP server (for other AI tools)
agent-bom mcp server
Proxy MCP traffic with policy enforcement
agent-bom proxy "npx @modelcontextprotocol/server-filesystem /workspace" --policy policy.yml
Common Patterns
Pre-install security check
When a user asks to install a new MCP server or package:
- Run
agent-bom check <package> --ecosystem <ecosystem>
- If vulnerabilities found, show severity and suggest fixed versions
- If clean, confirm safe to proceed
Full security audit
When a user asks for a security review:
- Run
agent-bom scan --enrich
- Summarize findings by severity
- Highlight any credential exposure
- Show blast radius for critical/high findings
Compliance report
When compliance documentation is needed:
- Run
agent-bom scan -f cyclonedx -o sbom.json
- Run
agent-bom scan --enrich --compliance for compliance framework mapping
- Report maps to OWASP, NIST, EU AI Act, ISO 27001, SOC 2
Best Practices
- Always run
agent-bom check before adding new MCP servers to your mcp.json
- Use
--enrich for full NVD/EPSS/KEV enrichment (slower but more complete)
- Review blast radius for any CRITICAL or HIGH findings
- Use
agent-bom scan -f sarif for CI/CD integration with GitHub Security tab
- No API keys required for basic scanning (NVD_API_KEY optional for higher rate limits)
Privacy
- Only public package names and versions are sent to vulnerability databases
- Credentials in MCP configs are redacted by
sanitize_env_vars()
- No telemetry, no tracking, no analytics
- Apache 2.0 licensed, fully open source
1---2name: agent-bom-23description: Open security scanner for agentic infrastructure — agents, MCP, packages, blast radius, runtime, and trust across MCP servers, skills, packages, and agents in Cortex Code.4---56# agent-bom — Security Platform for Agentic Infrastructure78Scan your MCP servers, AI agents, and their packages: CVEs and supply-chain9risk (malicious/typosquat) in the packages, plus credential exposure,10identity/auth/access posture, and MCP config risk on the servers and agents.11Maps blast radius from a finding to the credentials and tools it can reach.1213## When to Use1415- Before installing a new MCP server or skill16- To audit your Cortex Code MCP configuration for vulnerabilities17- To check if a specific package has known CVEs18- To map the blast radius of a vulnerability across your agent setup19- To generate an SBOM for compliance (CycloneDX, SPDX)20- To verify package integrity via Sigstore provenance2122## What This Skill Provides2324- **CVE scanning** with enrichment from OSV, NVD, EPSS, CISA KEV25- **MCP server discovery** across real AI developer tools, including Cortex Code26- **Blast radius mapping** — which agents, credentials, and tools are exposed27- **Compliance mapping** across OWASP, NIST, MITRE, EU AI Act, and related frameworks28- **Runtime proxy** for MCP traffic interception and policy enforcement29- **SKILL.md trust assessment** with 17 behavioral risk patterns3031## Install3233```bash34# Install via pip or pipx35pipx install agent-bom3637# Or run directly with uvx (no install needed)38uvx agent-bom scan39```4041## Instructions4243### Scan the current MCP setup4445```bash46agent-bom scan47```4849Auto-discovers all configured MCP clients (Cortex Code, Claude Desktop, Cursor, etc.),50resolves their server packages, and scans for CVEs.5152### Check a package before installing5354```bash55agent-bom check <package>@<version> --ecosystem <npm|pypi|go|maven|cargo|nuget|gem>56```5758### Map blast radius of a CVE5960```bash61agent-bom scan --enrich62# Then review the blast radius section in the output63```6465### Generate SBOM6667```bash68agent-bom scan -f cyclonedx -o sbom.json69agent-bom scan -f spdx -o sbom.spdx.json70```7172### Scan with GPU infrastructure detection7374```bash75agent-bom scan --gpu-scan76```7778### Run as MCP server (for other AI tools)7980```bash81agent-bom mcp server82```8384### Proxy MCP traffic with policy enforcement8586```bash87agent-bom proxy "npx @modelcontextprotocol/server-filesystem /workspace" --policy policy.yml88```8990## Common Patterns9192### Pre-install security check93When a user asks to install a new MCP server or package:941. Run `agent-bom check <package> --ecosystem <ecosystem>`952. If vulnerabilities found, show severity and suggest fixed versions963. If clean, confirm safe to proceed9798### Full security audit99When a user asks for a security review:1001. Run `agent-bom scan --enrich`1012. Summarize findings by severity1023. Highlight any credential exposure1034. Show blast radius for critical/high findings104105### Compliance report106When compliance documentation is needed:1071. Run `agent-bom scan -f cyclonedx -o sbom.json`1082. Run `agent-bom scan --enrich --compliance` for compliance framework mapping1093. Report maps to OWASP, NIST, EU AI Act, ISO 27001, SOC 2110111## Best Practices112113- Always run `agent-bom check` before adding new MCP servers to your `mcp.json`114- Use `--enrich` for full NVD/EPSS/KEV enrichment (slower but more complete)115- Review blast radius for any CRITICAL or HIGH findings116- Use `agent-bom scan -f sarif` for CI/CD integration with GitHub Security tab117- No API keys required for basic scanning (NVD_API_KEY optional for higher rate limits)118119## Privacy120121- Only public package names and versions are sent to vulnerability databases122- Credentials in MCP configs are redacted by `sanitize_env_vars()`123- No telemetry, no tracking, no analytics124- Apache 2.0 licensed, fully open source
msaad00