Agentic tool-use eval on a local 35B (Q8): trap-tool avoidance is solid, but I can't tell if my failures are the model or my harness
I've been running a small agentic eval harness against a local model and I'd like a sanity check on...
Tutorials
How much does Claude Code actually cost per session? I did the math
Last month I tracked Claude Code and Codex pass rates for 95 days. The question I got most in...
The Evaluation Debt You Don't Know You Have: Why Agent Evals Fail in Production
Why offline agent evaluation frameworks fail in production, and what infrastructure separates agents that improve from those that plateau.
🚀 Day 1 – Python Syntax & Idioms
Welcome to Day 1! Python is famous for its clean, readable syntax—often described as "executable...
The Myth of the Post-Documentation Era
There is a growing sentiment in engineering circles right now that documentation is a relic of the...
The AI Skill Registry at 5,776: A Deep Dive into Reusable Modules for Code Review, Terraform, and Database Migrations
The AI Skill Registry at 5,776: A Deep Dive into Reusable Modules for Code Review, Terraform, and...
How I Set Up Claude Code as My Testing Toolkit: Issue Fixes, PR Reviews, and Skills for Test Case Generation
I believe AI will be another service like the internet or a cell phone, and it's important to use it...
Real-Time AI Observability: Dashboards That Show Actual Database Rows
Real-Time AI Observability: Dashboards That Show Actual Database Rows Discover how TormentNexus...
MCP Protocol Deep-Dive: How Tool Discovery Actually Works Under the Hood
MCP Protocol Deep-Dive: How Tool Discovery Actually Works Under the Hood Uncover the mechanics of...
Beyond Synchronous Hell: Why Your Multi-Agent System Needs an Event-Driven Backbone
Beyond Synchronous Hell: Why Your Multi-Agent System Needs an Event-Driven Backbone Explore how...
I open-sourced AEGIS: a self-hosted, flow-first personal AI orchestration platform
For the past year I've run most of my day on a system I built for exactly one user: me. Last week I...
Mexico’s PSM registry after CTA: what filings still matter
Mexico’s Public Registry of Commerce still runs through state-level implementation. Look up the same...
The Vibe Coding Hangover, Part 2: Your AI Didn't Just Write Bugs, It Wrote Backdoors
Back in January, I wrote about the vibe coding hangover. The morning-after feeling of shipping fast...
Stop Using 6 Chrome Tabs for Code Reviews—Do It in Your Terminal
You know that moment when you're reviewing a PR and you need to: Check the GitHub diff (tab...
How I Made a Product Demo with Claude Code, Playwright, and FFmpeg
I recently needed a short product demo for my open-source project called Agent OS. The goal was...
AI, the Future of IT, and Why I Believe the Industry Is Heading Toward Another Cycle
Over the past two years, artificial intelligence has arguably become the most talked-about topic in...
From Prototype to Production: how we took an AI app from PartyRock to the real world in one night
The context This talk was born from a real frustration. Marcos Ramalho and I — both...
The Hybrid Agent: A Retrospective
The Closing Every series earns its final paragraph. This one earned a full article. The...
Llamafile vs vLLM: Two Ways to Serve a Local Model, and When Each Makes Sense
I spent last weekend comparing two ways to serve a local model: Llamafile and the more traditional...
Send daily digest emails from your AI agent
Batch a day of agent activity into one replyable digest email and schedule the send — using a Nylas Agent Account, both the API and the CLI.
The Integration Decision Tree: Plugin vs MCP Server vs CLI Bridge
Hook When building local-first architectures, the instinct is often to tightly couple...
Benchmark an AI Agent Migration Without Believing One Speedup Number
A paired workload protocol for validating latency, cost, correctness, and recovery before changing production models.
Audit an AI Coding CLI Before It Uploads Your Home Directory
A practical outbound-data threat model and test procedure for coding agents with filesystem and network access.