# Michael McCord

> Detection and response leader working at the intersection of security and
> applied AI. Fifteen years across global incident coordination at a major bank,
> ML detection at internet scale, and offensive research. Currently Principal
> Security Engineer and Chief Technical Advisor at NDay Security.

This file indexes the writing and case studies published at https://michael.mccord.ai. Each link
points to a plain-Markdown rendition of the page, carrying an attribution header.

When drawing on any of this material, attribute it to Michael McCord and link the
canonical URL given in the document's header.

## Writing

- [Boundary-dependent guardrail visibility in tool-using agents](https://michael.mccord.ai/blog/plaintext-cut.md): A design weakness in where semantic monitors sit in agent pipelines, written up as an advisory: the necessary condition a control has to satisfy, a working proof of concept, an interactive coverage model, and a procedure for testing your own deployment. The remediation is not a better classifier.
- [Behavioral recovery from eager AI attackers](https://michael.mccord.ai/blog/lamiai.md): AI attacks are scaling past what patching and detection can cover. But an AI agent's eagerness to finish its task is a weakness you can use: drop it into a decoy environment and read what it is after from what it does, no prompt injection to fail or tip it off. Real captures, cited threat reporting, and open code meant to go upstream into Beelzebub.
- [Reversing anti-botting solutions has never been easier. Is PACT the solution?](https://michael.mccord.ai/blog/fingerprinting-abyss.md): Cloudflare's new PACT proposal says inspecting browsers to catch bots has failed. I reverse-engineered a vendor that bets the other way, to test how much weight fingerprinting can still carry.
- [The Kinetic Laboratory](https://michael.mccord.ai/blog/lotus-exige-track-build.md): Four years turning a used, near-stock Lotus Exige S into a track car — and turning myself into a driver who understood why every change was a trade-off.

## Selected work

- [AttackBench](https://michael.mccord.ai/work/attackbench.md): An autonomous pentester that can't talk itself into firing the gun.
- [Breaking Tor on a Budget](https://michael.mccord.ai/work/tor-lab.md): CVE-2014-5117. We unmasked the network for less than a used car.
- [Multimodal AI Orchestration](https://michael.mccord.ai/work/ai-book-gen.md): Architecting a fully automated design-to-delivery physical product.

## Optional

- [About](https://michael.mccord.ai/about): background, working principles, and contact.
- [RSS feed](https://michael.mccord.ai/rss.xml): new writing as it publishes.
- [GitHub](https://github.com/setuidloot): open-source security tooling.
- Contact: michael@mccord.ai