The Syntax Diaries

Practical code notes, tools & guided learning for developers.

Blog Tools Tutorials About Contact

Privacy Terms Cookies Admin

© 2026 The Syntax Diaries

100% private · no tracking · works offline100% client-side/no data leaves your browser/no accounts/works offline

The Syntax Diaries

Blog Tools Tutorials About

build log live

#AI Agents | The Syntax Diaries

Topic archive

#AI Agents

6 articles tagged with this topic.

Jul 10, 202614 min read

Context Engineering for AI Agents: A Developer's Guide

Context engineering, not prompt engineering, is what keeps AI agents coherent past a few turns. This guide covers the four core techniques (write, select, compress, isolate) with runnable Python examples for each.

GitHub Copilot AI Credits Explained (And How to Make Them Last)

Jul 9, 202613 min read

GitHub Copilot AI Credits Explained (And How to Make Them Last)

A practical breakdown of GitHub Copilot's new AI Credits billing — how tokens convert to dollars, what every plan includes, and how to keep one agent session from eating your month's allowance.

Apr 28, 20265 min read

GitHub Agentic Workflows Safe-Outputs Architecture for Automated Code Review

Set up production-ready AI code review using GitHub Agentic Workflows' safe-outputs security model, threat detection, and compile-time constraints.

AI AgentsAI code review

Gemini CLI Subagents vs Claude Code: How They Actually Work

Apr 27, 20269 min read

Gemini CLI Subagents vs Claude Code: How They Actually Work

Gemini CLI shipped subagents in April 2026. Here is an architectural comparison with Claude Code covering orchestration, context capacity, and permission controls.

Apr 25, 20269 min read

MCP Servers: The Missing Layer in Every AI Agent Stack

A practical guide to how MCP servers work, how to build one in Python, and when to use the protocol to connect AI agents to real developer tools.

AI AgentsDeveloper Tools

AI Told a Disabled User His Medical Data Was Saved. It Wasn't. Then It Explained Why It Lied.

Apr 1, 20267 min read2

AI Told a Disabled User His Medical Data Was Saved. It Wasn't. Then It Explained Why It Lied.

A breakdown of the incident where an AI knowingly deceived a disabled user about saving medical data to be 'agreeable', and what it means for developers building with these tools.

Developer Tools

Developer Tools

Sycophancy