Skip to content

Gabo Rodriguez

role
Application Security · DevSecOps · Pentesting · AI Security
location
Buenos Aires, AR
github
github.com/y3rb1t4
linkedin
linkedin.com/in/grodriguezsec
cv
read · download pdf

0x0 overview

Application security and DevSecOps engineer with a pentester's background. I build security into the delivery pipeline — threat modeling, SAST and DAST, secure SDLC — and I have spent years on the offensive side, testing web, mobile and cloud targets. On the cloud side I work with IAM across AWS and Azure, leaning on short-lived credentials instead of static secrets so access stays auditable.

Most recently I have been working on the security of AI agent systems: colmena, my open-source trust firewall, applies deterministic per-agent policy to every tool call a coding agent makes, with no model in the hot path.

0x1 selected work

  • colmena

    2026

    A deterministic trust firewall for multi-agent AI coding. Every tool call an agent makes is evaluated against compiled YAML policy in under 15ms, with no model in the hot path and an append-only audit trail behind it.

    [Rust] [MCP] [AI security] [Policy as code]

    source →

  • LexMesh

    2026

    A multi-agent system for criminal defense work, built on LangGraph. A supervisor routes case material to specialist agents — case analysis, legal research, procedural strategy, document drafting — each with its own prompt, tools and slice of shared state.

    [Python] [LangGraph] [Anthropic API] [Pydantic]

    [private — source not public]