Source: https://toddschiller.com/blog/index.html
# Blog Posts
[](https://toddschiller.com/feeds/all.atom.xml "Blog RSS Feed")
New here? [Start with the curated set](https://toddschiller.com/pages/start-here.html), or grab the [RSS feed](https://toddschiller.com/feeds/all.atom.xml) to keep up with new posts. Shorter pieces are also collected on the [Notes](https://toddschiller.com/notes/index.html) page.
[Todd Schiller](https://toddschiller.com/pages/about.html)
- Note What size model does real-time on-device protection need?
Oct 14, 2026 System One decision models fit on-device behavior detection, and Chrome is building one into the browser. For real-time protection, small models trained for the task still win on latency and memory.
[Read more: What size model does real-time on-device protection need?](https://toddschiller.com/blog/on-device-slm-behavior-detection.html "Permalink to What size model does real-time on-device protection need?")
- Contact Center Fraud: highlights from my Cyber Crime Junkies episode
Aug 20, 2026 Highlights from my conversation with David Dean Mauro on Cyber Crime Junkies about contact center fraud, insider risk, AI agents, and locking down shadow AI.
[Read more: Contact Center Fraud: highlights from my Cyber Crime Junkies episode](https://toddschiller.com/blog/cyber-crime-junkies-highlights.html "Permalink to Contact Center Fraud: highlights from my Cyber Crime Junkies episode")
- Note Agentic Software Factories: Parallel agents diverge. Plan to converge.
Aug 12, 2026 Running coding agents in parallel is closer to a team of different developers than one developer multitasking, so plan a deliberate converge phase after the diverge phase.
[Read more: Agentic Software Factories: Parallel agents diverge. Plan to converge.](https://toddschiller.com/blog/parallel-agents-diverge-converge.html "Permalink to Agentic Software Factories: Parallel agents diverge. Plan to converge.")
- Note Loop Engineering: always be compounding
Jul 07, 2026 Loop engineering for AI agents should raise the floor on each iteration by creating tools and correcting instructions.
[Read more: Loop Engineering: always be compounding](https://toddschiller.com/blog/loop-engineering-compounding.html "Permalink to Loop Engineering: always be compounding")
- Note Beliefs hitchhike like genes
Jun 09, 2026 A new explorer that clusters public thinkers by the genome of their beliefs, inspired by genetic hitchhiking.
[Read more: Beliefs hitchhike like genes](https://toddschiller.com/blog/belief-phylogenetics-artifact.html "Permalink to Beliefs hitchhike like genes")
- Note agent-browser-shield is live on Product Hunt
Jun 05, 2026 agent-browser-shield, a source-available defense layer for AI agents in the browser, is on Product Hunt today.
[Read more: agent-browser-shield is live on Product Hunt](https://toddschiller.com/blog/agent-browser-shield-product-hunt.html "Permalink to agent-browser-shield is live on Product Hunt")
- Note agent-browser-shield June 3 update: 14 new rules and the Chrome Web Store listing
Jun 03, 2026 14 new rules shipped in agent-browser-shield, now installable from the Chrome Web Store, and using it for daily driving.
[Read more: agent-browser-shield June 3 update: 14 new rules and the Chrome Web Store listing](https://toddschiller.com/blog/agent-browser-shield-june-03-update.html "Permalink to agent-browser-shield June 3 update: 14 new rules and the Chrome Web Store listing")
- Note Introducing agent-browser-shield (alpha): keeping AI agents safe in the browser
Jun 01, 2026 Announcing the alpha of agent-browser-shield — a source-available defense layer (also available as an OpenClaw skill) that blocks prompt injection and dark patterns before they reach your agent.
[Read more: Introducing agent-browser-shield (alpha): keeping AI agents safe in the browser](https://toddschiller.com/blog/agent-browser-shield-alpha.html "Permalink to Introducing agent-browser-shield (alpha): keeping AI agents safe in the browser")
- Note Letting OpenClaw loose on Boston's open data
May 30, 2026 At the Boston OpenClaw 2026 hackathon, I let an agent autonomously connect to the city's open-data MCP server, devise its own corruption-signal queries on contract data, and package the workflow...
[Read more: Letting OpenClaw loose on Boston's open data](https://toddschiller.com/blog/openclaw-boston-open-data.html "Permalink to Letting OpenClaw loose on Boston's open data")
- Note Did Google sneak a local LLM model into Chrome?
May 26, 2026 A response to FUD around Chrome's new Local AI models, plus a demo using PixieBrix + Local AI to filter my LinkedIn feed.
[Read more: Did Google sneak a local LLM model into Chrome?](https://toddschiller.com/blog/chrome-local-ai-linkedin-filter.html "Permalink to Did Google sneak a local LLM model into Chrome?")
- [Newer](https://toddschiller.com/blog/index.html#)
- [1](https://toddschiller.com/blog/index.html)
- [2](https://toddschiller.com/blog/index2.html)
- [3](https://toddschiller.com/blog/index3.html)
- [4](https://toddschiller.com/blog/index4.html)
- [Older](https://toddschiller.com/blog/index2.html)