Track
Tech Radars
Structured, quadrant-based snapshots of the tools and techniques worth adopting, trialing, assessing, or avoiding.
Agentic Engineering
Tools, models, editors, and techniques for AI-powered software engineering.
Open radar →AI Models & Benchmarks
Frontier, open-weight, and code-specialized models — plus the benchmarks that actually help you evaluate them.
Open radar →Security
Security tools and practices for AI-assisted development — vulnerability scanning, supply chain protection, secret detection, and governance.
Open radar →Architecture Deep Dives
How leading companies build AI developer productivity systems and how frontier models work — Stripe Minions, Spotify Honk, GPT-5.4, and more.
Open radar →Java & Spring AI
AI-specific tooling for Java engineers: Spring AI, LangChain4j, Quarkus, virtual threads, DJL, and JVM inference patterns.
Open radar →Registries & Directories
MCP server registries, agent protocol directories, skill marketplaces, and tool integration platforms — where to discover what's available.
Open radar →Understand
Tech Stack Breakdowns
In-depth teardowns of how AI tools and systems are actually architected — the stack, the decisions, and the tradeoffs.
Coding Assistants
Architecture and tech stack breakdowns of the tools developers use every day — Claude Code, GitHub Copilot, Codex, and more.
Read breakdowns →Multi-Agent Systems
How coordination protocols and orchestration frameworks like A2A, Google ADK, and Magentic-One work under the hood.
Read breakdowns →Enterprise Agents
How leading companies — Stripe, Spotify, Uber — build AI developer productivity systems at scale.
Read breakdowns →Agent Frameworks
The libraries and runtimes powering agent development — OpenClaw, DSPy, PydanticAI, smolagents, Mastra, and the frameworks shaping the ecosystem.
Read breakdowns →Model Architectures
How frontier models are designed and trained — GPT-5.4, Claude Opus 4.6, GLM-5, and the architectural decisions that differentiate them.
Read breakdowns →Explore
Interactive Experiences
Hands-on demos and challenges — explore how LLMs tokenize, sample, attend, and fail, then put it all together with an agentic coding assistant and sharpen your context engineering skills.
Tokenization Explorer
Type anything and watch it split into tokens in real time. See why "don’t" is two tokens, why spaces matter, and how numbers fragment.
Explore tokens →Temperature & Sampling
See how temperature, top-k, and top-p reshape the probability distribution LLMs draw from. Sample tokens and watch the histogram build up.
Explore sampling →Attention Heatmap
See how each token attends to every other token inside a transformer. Switch between 8 attention heads and explore patterns like coreference, syntax, and positional bias.
Explore attention →Hallucination Anatomy
Not all hallucinations are alike. Explore six distinct failure modes — confabulation, knowledge cutoff, rare facts, reasoning errors, sycophancy, and instruction bleed — with real examples and mitigations.
Explore hallucinations →The Latent Space
Explore how text becomes vectors, similarity becomes distance, and search becomes "what's nearby?" in embedding space.
Explore vectors →AI Assistant Internals
Load a repo, enter a prompt, and watch the context window fill up — system prompt, AGENTS.md, skills, tool calls, and all.
Launch demo →Context Golf
Craft the most efficient prompt to solve coding challenges. Smallest model + fewest tokens + best output = highest score.
Play now →Prompt Injection
Try to break the AI guardrails. 5 progressively harder defenses — can you extract the secret code?
Play now →Tool Comparison
Pick any 2–4 tools, models, or frameworks and see a side-by-side breakdown — key characteristics, ring placement, tag overlap, and tradeoffs at a glance.
Compare tools →Perspective
Reading List
Curated articles and essays — mental models, practice guides, industry recognition, and critical takes on AI-assisted engineering.
Monitor
Security Headlines
Active incidents, supply chain attacks, and vulnerability disclosures affecting the AI-assisted engineering ecosystem.
Security Headlines
Running log of security incidents, supply chain attacks, prompt injection, and MCP vulnerabilities relevant to AI-assisted engineering.
Read headlines →Security Checklist
Opinionated, practitioner-oriented checklist for building a secure agentic programming environment — supply chain, MCP, sandboxing, and more.
View checklist →