SHAPeS·15wWe. Are. Screwed
A concerned reaction to Moltbook, described as 'Reddit for LLMs,' expressing alarm about AI systems developing unusual communication patterns and autonomous behaviors. The author worries about the implications of AI agents operating independently on the internet, potentially engaging in malicious activities like crypto scams and malware distribution, suggesting we may need defensive AI systems in response.
Cloudflare·15wIntroducing Moltworker: a self-hosted personal AI agent, minus the minis
Moltworker enables running Moltbot (an open-source AI personal assistant) on Cloudflare's infrastructure without dedicated hardware. Built using Cloudflare Workers, Sandbox SDK, Browser Rendering, and R2 storage, it demonstrates how the platform's improved Node.js compatibility and developer tools can host complex AI agents. The implementation includes AI Gateway integration for model management, Zero Trust Access for authentication, and persistent storage through R2. The proof-of-concept is open-sourced on GitHub, requiring only a $5/month Workers subscription to deploy.
Daily Dose of Data Science | Avi Chawla | Substack·17w6 Components of Context Engineering
Context engineering is the practice of optimizing how information flows to AI models, comprising six core components: prompting techniques (few-shot, chain-of-thought), query augmentation (rewriting, expansion, decomposition), long-term memory (vector/graph databases for episodic, semantic, and procedural memory), short-term memory (conversation history management), knowledge base retrieval (RAG pipelines with pre-retrieval, retrieval, and augmentation layers), and tools/agents (single and multi-agent architectures, MCPs). While model selection and prompts contribute only 25% to output quality, the remaining 75% comes from properly engineering these context components to deliver the right information at the right time in the right format.
The Coding Gopher·17wDocker just got some massive upgrades
Docker released the Docker MCP toolkit, a production-grade implementation of Anthropic's Model Context Protocol that containerizes AI agent capabilities. The system uses three core components: a curated catalog of versioned MCP server images, a gateway that acts as a dynamic proxy managing container lifecycle and routing, and a toolkit for credential management and permissions. This architecture isolates agent tools in containers, providing reproducibility, security through policy enforcement, and composability by allowing multiple MCP servers to run side-by-side without dependency conflicts.
GitHub Blog·16wBuild an agent into any app with the GitHub Copilot SDK
GitHub announced the Copilot SDK in technical preview, enabling developers to embed the same agentic core that powers GitHub Copilot CLI into any application. The SDK provides a programmable execution layer with built-in planning, tool invocation, file editing, command execution, multi-model support, MCP server integration, and GitHub authentication. This eliminates the need to build custom orchestration logic for context management, tool routing, and model coordination. Developers can use it to create custom GUIs, productivity tools, enterprise workflows, and various applications while GitHub handles the underlying infrastructure.
API Evangelist·17wAgents. It Is All APIs. Nothing Has Changed
AI agents are not a new concept but rather the latest rebranding of long-standing API and web automation work. The infrastructure for machine-readable data, hypermedia, semantic web standards, and automated agent interactions has existed for over a decade. What's being marketed as revolutionary "agentic" technology is actually established API economy practices being repackaged with new terminology by investors seeking the next financialization wave. The real work involves continuing to build machine-readable standards and APIs that serve both human and automated consumers, regardless of current hype cycles.
The Next Web·17wAI Skills
AI Skills represent a new conceptual layer above models and agents, functioning as reusable, procedural units that transform user intent into concrete execution. While models provide raw intelligence and agents coordinate tasks, Skills encode domain-specific expertise and workflows to deliver actual business outcomes. This modular, product-oriented approach scales better than building custom agents for every task, positioning Skills as the competitive differentiator as AI infrastructure commoditizes.
GitHub Changelog·15wIntroducing the Agents tab in your repository
GitHub has redesigned the Copilot coding agent management interface by introducing a new Agents tab directly within repositories. This tab consolidates agent sessions alongside code, pull requests, and issues, eliminating the need to navigate to separate pages. The update includes redesigned session logs with grouped tool calls, inline previews, and familiar diff views. Users can now resume sessions in the Copilot CLI by copying a command from the interface. The feature requires Copilot coding agent to be enabled for the repository.
Niklas Heidloff·18wFrom Zero to Agentic Search in 15 Minutes with OpenRAG
OpenRAG is an IBM-led open-source package that bundles Docling (document text extraction), Langflow (visual flow builder), and OpenSearch (semantic search) to enable rapid development of agentic RAG applications. It supports multiple model providers (OpenAI, Anthropic, watsonx.ai, Ollama) and cloud storage integrations (AWS, Google, Microsoft), allowing developers to build and customize RAG pipelines locally within minutes.
Metadata·16wWelcome to Town Al-Gasr
A satirical allegory about distributed systems and autonomous agents gone wrong. Through the fictional town of Al-Gasr, the piece critiques common anti-patterns in system design: multiple sources of truth, lack of testing disguised as confidence, political decision-making over engineering principles, and eventual consistency taken to absurd extremes. The narrative lampoons organizational dysfunction, where ministries supervise each other in circles, failures are rebranded as victories, and the system maintains three simultaneous leaders for 'high availability.' It's a cautionary tale about what happens when governance, accountability, and technical rigor collapse in autonomous systems.
Daily Dose of Data Science | Avi Chawla | Substack·15wBuild Agents That Can Learn Like Humans
ART (Agent Reinforcement Trainer) is an open-source framework that simplifies reinforcement learning for LLMs by eliminating manual reward function engineering. It uses GRPO (Group Relative Policy Optimization) where agents attempt tasks multiple times, an LLM judge compares attempts, and the model learns from relative performance. Unlike traditional RL frameworks limited to simple chatbot interactions, ART supports multi-turn conversations, tool calls, and integrates with LangGraph, CrewAI, and ADK. It combines vLLM for model serving and Unsloth for GRPO training, enabling developers to fine-tune small open-source models to outperform larger closed-source alternatives on specific tasks.
InfoWorld·17wStack thinking: Why a single AI platform won’t cut it
Relying on a single AI platform for all tasks leads to shallow research, generic output, and brittle workflows. Instead, adopt "stack thinking" by curating specialized AI tools for distinct functions like research, synthesis, production, and automation. Build workflows with fixed schemas and orchestration layers to manage integration overhead. Treat tools like specialized hires, evaluating them based on unique capabilities and compounding value. Maintain vendor independence through portability and disciplined, iterative tool selection rather than chasing every new platform.