Judgment is the new bottleneck.
Engineering patterns build the practice.
A curated library of prompts and agentic patterns from professional sources, articles from practice, and a learning system in early access.
Skill: /init CLAUDE.md and skill setup
A comprehensive onboarding flow for setting up CLAUDE.md and related skills/hooks in the current repository, including codebase exploration, user interviews, and iterative proposal refinement.
What working prompts look like.
Every prompt comes from a working codebase or a published pack, classified across five axes so you can find the shape you need. 1,000+ are free to read in full.
Documentation Access
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository. Documentation Access. When working with external libraries or frameworks, use the Context7 MCP to...
Skill Scanner
Scan agent skills for security issues before adoption. Detects prompt injection, malicious code, excessive permissions, secret exposure, and supply chain risks.
Migration and code evolution instructions generator for GitHub Copilot
Migration and code evolution instructions generator for GitHub Copilot. Analyzes differences between two project versions (branches, commits, or releases) to create precise instructions allowing Copil
Or point your AI at it.
Hand your assistant one priming prompt and it can survey over 13k classified prompts, pull the patterns that fit your stack, and turn them into Skills and workflows for your own setup.
$ curl "https://mlad.ai/api/v1/prompts?theme=debugging&tech=python" \
-H "Authorization: Bearer mlad_live_…"
{ "prompts": [
{ "id": "prompt-4e51e0f6bf5b", "promptType": "System",
"activation": "Persistent", "techStack": ["python"], … } ],
"total": …, "hasMore": true }
Keeping AI on track is an art.
go deeper with your grounding, but don't drink the cool-aid... stay critical. give me some ideas/options
Models are improving rapidly, but we've yet to see the stand-alone miracles that hype promises. Since its inception, AI-coding and Agentic-development have required a lot of craft, filtering, and smart work-patterns to get reliable results.
Whether you're looking for more effective ways of working, or how to increase the value you can derive from modern software-engineering practices, MLAD is here to share the best of what we've found. So that your team can build reliable workflows and better systems.
Articles for AI Developers.
![[contain] The four phases of the proposal-driven cycle](/article-images/2026/01/proposal-cycle-simple-wm-1.png)
The Proposal-Driven Cycle
A lightweight workflow for AI coding that captures intent alongside code.
Skills Are More Context-Efficient Than MCP
MCP servers can consume 55,000 tokens of Claude Code's context window before you type.
Securing Your Vibe-Coded App
Nearly half of AI-generated code contains security flaws.
The learning system, in early access.
Quests built on recorded AI sessions. Watch an engineer work with AI, commit to what you'd do at the judgment moments, then see the gap. The first quest is live.
Follow the build. Predict. Commit. See the gap.
Reddit Discovery & Analysis
From blank repo to 3,310 scored signals, 42 clusters, and interactive hierarchy visualization.
Build a complete Reddit analysis pipeline using AI-assisted development. Watch real engineering sessions, answer challenges at key moments, build judgment across six pattern families.
Cognitive effort — and even getting painfully stuck — is likely important for fostering mastery.
Anthropic 2025 RCT
Dopamine neurons don't just respond to any old reward: they respond only to rewards that differ from their prediction.
Schultz, 2016
Build your skills with every judgment call. Evidence fades between sessions. Practice restores it.
summarized on your Skills Profile