|4 min read
What Belongs in Agent Memory
A durable handoff separates observations, hypotheses and unfinished work, then records what would make each observation stale.
[AI & Data][Software Engineering]
55 posts
A durable handoff separates observations, hypotheses and unfinished work, then records what would make each observation stale.
The $7.99 rebuild bought a first version; the repository now shows the less photogenic work of preserving an archive and keeping its promises.
A review bundle should connect the requirement to evidence for the exact candidate and make the remaining decision explicit.
A dedicated email address limits which messages reach an agent; the harder details are untrusted content, sending authority and missing attachments.
Tracing a WHOOP event through receipt, diagnostic capture, Sheets forwarding and reconciliation reveals what a successful response leaves unproved.
A deliberately broken discount function passes its edited test; an unchanged acceptance check shows why the owner of the verdict matters.
A useful agent acts within delegated authority, prepares consequential actions for review, and asks again when the action changes.
Stop prompting, start looping — but the loop is the easy 20%. After a year shipping agentic loops (and a $500 runaway), here's the part the hype keeps getting backwards: the done-check and the verifier.
One paper shows frontier models degrade as context grows — even on trivial tasks. The other shows reasoning models hit a wall and think less as problems get harder. Read carefully, both point at the same engineering response.
Principles 26–30. The calibration layer that catches what the rest of the framework would miss: a PR-noise budget, independent verification, model-swap regression discipline, the 15-tool-call rule, and protecting junior development.
Principles 15–20. The harness configuration that keeps the kernel and lifecycle cheap: CLAUDE.md under 200 lines, hooks for real incidents, skills that auto-invoke, subagent isolation, pinning, and Stage 5 distribution.
Principles 6–14. How work moves through an agentic engineering team: the ticket as contract, AI distillation with human curation, three gates, verification before done, characterisation tests, the 1.2× capacity rule, the J-curve, and telemetry.
Principles 1–5. The five rules that everything else in the framework rests on: standardise the harness, make verification load-bearing, default to plan mode, pick the cheapest layer, reflect every task.
Practitioner consensus puts the cliff around fifteen tool calls per prompt. Here's why agents degrade past that, and the three operational rules that keep them on the safe side.
Anthropic's multi-agent Research feature beat single-agent Opus 4 by 90.2% — at 15× the token cost. Every documented production swarm runs on rails. Here's the topology decision framework before you commit.
Agents over-refactor stable code without a safety net. Feathers' characterisation-test technique — write tests for current behaviour before changing anything — is more important than ever. The agent itself is the perfect characterisation-test-writer.
Karpathy named one mode. Willison named the other. Most 'AI failed in production' stories are actually 'we promoted a vibe-coded prototype without transitioning into the production discipline.'
ReAct gave us a three-step loop. Production hardened it into five. The two new steps — Plan and Verify — are where everything that goes wrong, goes wrong. And the field has now named the worst offender.
Three open-source extractions converged on the same five layers. The architecture isn't a vendor narrative — it's a discovered structure. Here's the decision rule that keeps you from over-engineering it.
Most teams plateau at Stage 2 because they confuse 'we built skills' with 'we have a working AI engineering culture.' Here's the 5-stage diagnostic — and the moves that get you from Individual to Distributed.
When I first built Gluon on my Mac mini, I was solving a personal problem: monitoring Claude agents without losing my mind to tmux logs. But when teams join the picture, everything changes — security, governance, observability, and the fundamental role of the developer. Here's what production infrastructure for autonomous agents looks like.
How Gluon's Ralph Loop enables autonomous Claude execution with built-in safety rails — circuit breakers, multi-signal completion detection, and cost controls that scale from simple tasks to complex workflows.
I built an open-source MCP server that reduces LLM token usage by 70-90% through server-side HTML filtering, markdown conversion, and CSS selector targeting. Here's why context efficiency matters—and how Scraper MCP solves it.
I gave Claude Code an XML backup of my 19-year-old WordPress blog and asked it to rebuild everything as a modern NextJS site. What happened next was like watching a swarm of expert developers work in parallel—spawning agents, debugging TypeScript errors, and shipping production-ready code. All in 26 minutes. For eight dollars.
After watching 40% of agentic AI deployments fail in production, I'm building Dagentic — a serverless-first framework designed for what AI agents actually are: unpredictable, spiky workloads that modify themselves mid-execution.
After 12 months of systematic optimization, I've documented 50-70% productivity gains with AI coding assistants. The secret isn't just using AI tools—it's teaching them to think like you do through carefully crafted configurations.
I built a multi-agent system that researches a topic and hands back a formatted Word document — citations, images, the lot — in minutes. Here's how the agents divide the work, and the one part the machine still can't own.
At 3 AM, I was manually cropping 47 personal photos for a LoRA model when I realized half were the wrong aspect ratio. Three hours wasted. So I built a simple Python app that does the same work in 15 minutes—and it changed how I think about AI tooling infrastructure.
Confused by Red Hat's Linux ecosystem? Learn how RHEL, Fedora, and CentOS evolved from one distribution into three distinct solutions for different needs.
Updated ZipFileInputFormat framework for processing thousands of ZIP files in Hadoop with failure tolerance and comprehensive examples
Automating Sea Turtle mount acquisition in World of Warcraft using custom waypoint navigation and fishing pool detection algorithms
Advanced navigation techniques for autonomous MMORPG characters using Recast/Detour navigation meshes and path finding algorithms
Build a Java utility class to consume Twitter Streaming API data for offline analysis in Hadoop with automatic file segmentation
Custom utility classes to extract and parse ZIP file contents in Hadoop MapReduce jobs using ZipFileInputFormat and ZipFileRecordReader
Quick shell command to clear all wiki pages from a freshly installed Trac environment using trac-admin
Enhance Apache's mod_evasive security module with detailed environment variable reporting similar to mod_security for better attack analysis
Deep technical analysis of how comment spam evolved from simple bots to sophisticated coordinated botnets, including real attack logs and countermeasures
I wrote this RC4/ARCFOUR implementation in PHP - based on the original C source code posted on usenet in 1994.
It may seem a little sad but I can honestly say that reading my accesslog is far more interesting than any soap opera on TV; they are filled with e...
Building a tool to detect and convert character encodings in the FreeDB database, processing 1.8 million entries with surprising results
Discover how to unlock your DVD writer's full potential by editing firmware burn speeds with this handy tool when media won't burn at rated speeds
Quick reference to Oracle TopLink version compatibility issues when importing older project files into newer workbench versions
I recently had to use the HTTP::Lite perl module and thought I would share an example of how to use it.
A quick and simple Velocimacro to generate HTML Select boxes and its Options; automatically selects the correct option based on the value given.
JSSHTunnel is a lightweight GUI application which forwards local & remote ports over SSH connections.
This is a very simple utility function to produce an indented XML document as a String given a Document.
Learn how type sensitivity in Velocity template comparisons can break your HTML forms and how to fix it with proper type conversion
If you're reading this article you've probably encounted the same problem as I have.
I was working on a web-based application recently and wanted a lazy way to get data from a simple 'value object' out to the front end without havin...
Essential reading for every developer: Joel Spolsky's definitive guide to Unicode and character encoding fundamentals
Builds a JAR file Test. jar and generates a manifest specifying that class Test is the Main-Class.
A better random number generator java import java. security.
When using JDK 1. 4 and above...
Returns true on success, false on failure: java public static boolean deleteDir(File dir) { if (dir.
I made this back in April 2004, it only took a couple of hours to write and build the database.