PantheonLM (formerly BMAD-CYBERSEC) is an enterprise-grade multi-agent orchestration framework purpose-built for cybersecurity operations. It ships with 81+ pre-built agents that cover the full security lifecycle.
# npm
npm install @blackunicorn/pantheonlm
# pnpm
pnpm add @blackunicorn/pantheonlm// pantheon.config.ts
import { defineConfig } from '@blackunicorn/pantheonlm'
export default defineConfig({
llmProvider: 'anthropic',
apiKey: process.env.ANTHROPIC_API_KEY,
agents: {
enabled: ['ThreatIntelAgent', 'VulnResearchAgent', 'ReportAgent'],
},
logging: {
level: 'info',
output: './logs/pantheon.log',
},
})import { PantheonLM } from '@blackunicorn/pantheonlm'
const pantheon = new PantheonLM()
const result = await pantheon.run('ThreatIntelAgent', {
target: 'CVE-2024-1234',
depth: 'full',
})
console.log(result.summary)
console.log(result.iocs) // Indicators of Compromise
console.log(result.mitigations)import { PantheonLM, Pipeline } from '@blackunicorn/pantheonlm'
const pantheon = new PantheonLM()
// Build a sequential pipeline
const pipeline = new Pipeline([
{ agent: 'VulnResearchAgent', input: { cve: 'CVE-2024-1234' } },
{ agent: 'ThreatIntelAgent', dependsOn: 'VulnResearchAgent' },
{ agent: 'ReportAgent', dependsOn: ['VulnResearchAgent', 'ThreatIntelAgent'] },
])
const report = await pantheon.execute(pipeline)
console.log(report.markdown)The full API reference including all 81+ agent signatures, configuration schema, and plugin development guide is available in the GitHub repository.