export default {
name: 'Miguel Ocampo',
alias: 'MiguelAutomate',
desc: 'breaking agents before they break your business',
type: 'human agent',
roles: ['AI red teamer', 'automation architect', 'security researcher', 'OSS author'],
currentFocus: ['agentic security', 'LLM red teaming'],
interests: ['neuroscience', 'JavaScript', 'mechanical engineering'],
education: 'B.S. Western Governors University',
alignment: 'lawful adversarial',
locale: {
current: 'La Jolla, California',
origin: 'San Diego, California'
},
methodology: 'TRACE → BREACH → IMPACT → PROOF',
proofOfWork: {
criticalFindings: 4,
openSourceProjects: 5
},
sideEffects: true
}Work
- ZeroSilence AI — autonomous vulnerability scanner for AI agent frameworks
- InjectionVault — Julia red-teaming framework, OWASP LLM Top 10 + MCP trust boundaries
- shadowClone — lock-free contention guard for shared JS state; scoped clones instead of blocking locks
- vortexVote — dependency-free threshold-gating rules engine
Portfolio · Book a call · Email
Fill in the repo URLs and add a line under Work for the two critical findings if you c