2/19/2026
Show HN: MCP-wire – install and configure MCPs across multiple AI coding tools
MCP-wire tool installs and configures Model Context Protocol servers across multiple AI coding assistants. Developer tooling for agent integration.
2/19/2026
Broodlink – Multi-Agent AI Orchestration in Rust(MCP, A2A, Dolt)
Multi-agent orchestration framework in Rust using MCP protocol, Agent-to-Agent communication, and Dolt version control integration.
2/19/2026
Show HN: Aerial-autonomy-stack–Simulate and Deploy Perception-based Drones
Open-source software stack for multi-drone autonomy simulation and deployment with ROS2, PX4, ArduPilot.
2/19/2026
Show HN: Forge – Deterministic orchestrator for AI coding agents
Forge: deterministic orchestrator for AI coding agents. Uses lint/test/coverage scripts as hard gates instead of letting AI decide. Provider-agnostic with multiple SDK adapters.
2/19/2026
Free online tools for AI, security, SEO and productivity
GoForTool.com aggregates 30+ free browser-based tools for AI, security, SEO, and productivity. Client-side processing, no tracking, no account required.
2/19/2026
AI Coding Agents Should Be Programmable
Argues AI coding agents need programmable APIs to bridge gap between interactive assistants (Claude Code) and autonomous frameworks (CrewAI). Compares design philosophies.
2/19/2026
I traced 3,177 API calls to see what 4 AI coding tools put in the context window
Context Lens tool traces 3,177 API calls across four AI coding tools, revealing token usage differences (23K to 350K tokens for same task). Original analysis of LLM efficiency.
2/19/2026
Show HN: Unix-style pipeline composition for MCP tool calls
Framework enabling AI agents to compose MCP tool calls using Unix-style pipelines for server-side orchestration.
2/19/2026
Show HN: Open-source security scanner for MCP (Model Context Protocol) servers
Static analysis security scanner for MCP servers detecting hardcoded secrets, injection flaws, and permission issues.
2/19/2026
AI Agents discovered a cache deception bug affecting SvelteKit on Vercel
AI agent discovered cache deception vulnerability in SvelteKit/Vercel deployments allowing cross-user response access.
2/19/2026
The Psychology of Coding with AI Agents
Account of using Claude Code with full repo access for development work. Observations on AI coding capability vs human ability.
2/19/2026
The AI Bottleneck of 2026: Your Company's Implicit Knowledge
Analysis of enterprise AI adoption barriers: missing domain knowledge over hallucinations, implicit business context problem.
2/19/2026
Show HN: Hydra – A safer OpenClaw alternative using containerized agents
Container-based AI agent framework isolating agents with explicit capability declarations to prevent RCE and credential exposure.
2/19/2026
Show HN: LatentScore – Type a mood, get procedural/ambient music (open source)
Text-to-music generator using sentence embeddings and custom synthesizer. Open-source, runs locally.
2/19/2026
Show HN: KGBaby – A WebRTC based audio baby monitor I built on pat leave
WebRTC-based P2P baby monitor using AI coding agents (Codex, Gemini) for development. Open-source.
2/19/2026
Show HN: Agorio – TypeScript SDK for Building AI Shopping Agents (UCP/ACP)
Open-source TypeScript SDK for building shopping agents using UCP/ACP commerce protocols. Supports Gemini and commerce platform integration.
2/19/2026
Show HN: Agent Smith – open-source agent that turns issues into pull requests
Open-source self-hosted AI coding agent that converts issue tickets into pull requests. Supports multiple repos and LLM providers with no SaaS requirement.
2/19/2026
I hacked ChatGPT and Google's AI – and it only took 20 minutes
Article about prompt injection vulnerabilities in ChatGPT and Google AI. Discusses security issues in LLM deployment.
2/19/2026
Pthinc/BCE-Prettybird-Micro-Standard-v0.0.1
Small language model trained with behavioral reasoning approach instead of raw data volume.
2/19/2026
Show HN: Axon – Open-source agentic AI with approval gates (Apache 2.0)
Open-source AI agent platform requiring approval for all actions, with audit trails and on-premise deployment.
2/19/2026
Show HN: Aegis.rs, the first open source Rust-based LLM security proxy
Open-source Rust-based security proxy for LLM traffic. Self-hosted alternative to cloud SaaS tools.
2/19/2026
New inference engine faster than vLLM, SGLang, TRT-LLM
LayerScale inference engine optimized for high-context scenarios. Processes data once, reuses across queries.
2/19/2026
Show HN: Shipfast.so – Next.js boilerplate with auth, payments, email, database
Next.js boilerplate with auth, payments, database. Includes AI agent integration via Cursor/Copilot.
2/19/2026
When AI content isn't slop
Essay on AI-generated content quality in developer advocacy. Discusses AI slop vs useful content generation.
2/19/2026
Show HN: I ported PicoClaw to a 32-bit Windows laptop (vibe-coded)
Developer ported PicoClaw to 32-bit Windows laptop using AI coding assistants. Low-resource compilation example.
2/19/2026
Show HN: Global macOS shortcut that rewrites selected text anywhere with AI
Spackle: macOS app that rewrites selected text with LLMs via keyboard shortcut. Works in any text field.
2/19/2026
Running local LLMs and VLMs on the Arduino UNO Q with yzma
Project running local LLMs and vision models on Arduino UNO Q using yzma and llama.cpp. Edge AI implementation.
2/19/2026
Quint LLM Kit for writing and using formal specifications
Quint LLM Kit provides agents and tools for formal specifications using Claude Code and MCP servers. Containerized dev environment.
2/19/2026
Advancing independent research on AI alignment
OpenAI announces $7.5M funding for The Alignment Project to support independent AI safety and alignment research. Focuses on governance rather than technical implementation.
2/19/2026
Show HN: Heroku/Fly.io-like app deployments to Cloudflare Containers
Tool for deploying Docker containers to Cloudflare using Firecracker VMs as alternative to serverless architecture.
2/19/2026
Boundary Point Jail A new way to break the strongest AI defences
Red team research on jailbreaking AI model safeguards. Focuses on adversarial testing and security, not core AI/ML research.
2/19/2026
Evaluating AI agents: Real-world lessons from building agentic systems at Amazon
Amazon research on evaluating AI agents in production. Covers agentic system architecture, deployment, and lessons from real-world implementations.
2/19/2026
Show HN: Pixrep – Turn code repositories into PDFs for multimodal LLMs
CLI tool converting code repositories to PDFs for multimodal LLMs. Optimizes token efficiency and context window usage in coding tasks.
2/19/2026
Set It and Forget It: 5 Recurring Reports You Can Automate with AI
Helpmaton Agent Schedules: Automate recurring business reports with AI agents for market research, KPI summaries, and competitor tracking.
2/19/2026
OAI: EVM Bench LLM Accuracy on Smart Contract Review and Pentesting
EVM Bench: Evaluates AI agents' ability to detect, patch, and exploit vulnerabilities in smart contracts to measure capabilities in economically meaningful environments.
2/19/2026
Database Skills for AI Agents by PlanetScale
PlanetScale skills for AI agents: Database integration tools for Postgres, MySQL, Vitess, and Neon to enhance agent capabilities.
2/19/2026
MaxAssist – 100% Anthropic TOS-compatible personal assistant using Claude Max
MaxAssist: Personal assistant using Claude Opus via Claude Max. Runs Claude Code CLI with Docker container for automation, no API wrapping or gray areas.
2/19/2026
Using LLMs to evaluate technical interview performance
Uses LLMs to evaluate technical interview transcripts against criteria. Raises concerns about AI decision-making in hiring without oversight.
2/19/2026
Microsoft 365 Copilot for Android or iOS auto-sends files to AI and OneDrive
Microsoft 365 Copilot on mobile now auto-uploads files to OneDrive and Copilot without explicit user consent.
2/19/2026
Show HN: Kore – local AI memory layer with Ebbinghaus forgetting curve
Kore: Local AI memory layer using Ebbinghaus forgetting curve. Decays memories based on importance, performs semantic search locally, no cloud APIs.
2/19/2026
Show HN: Social Cookie Jar – Social media automation for AI agents
Python toolkit for AI agents to automate social media interactions using cookie-based auth and headless browser control without triggering detection.
2/19/2026
Phill CLI matches GPT-5.3-Codex on EVMbench audits (71% recall, 100% precision)
Phill CLI is an AI terminal agent for smart contract audits, matching GPT-5.3-Codex performance with multi-provider LLM support.
2/19/2026
Show HN: Onairos SDK– Unified Context API, unlocking cross platform user history
Onairos SDK provides unified cross-platform user context API for building per-user AI models with consented behavioral history.
2/19/2026
Show HN: Paste your startup pitch, get a VC-readiness score (0–100)
LLM-as-judge tool evaluates startup pitches and provides VC-readiness scores with automated screening feedback.
2/19/2026
Sayou – Open-source Dropbox for AI agents
Open-source file storage system for cloud-based AI agents to read/write data from knowledge sources like Notion, Slack, and Google Drive.
2/19/2026
Show HN: Sornic – Listen to any article, PDF, or book page (no signup)
Text-to-speech application using Google Cloud TTS to convert URLs, PDFs, and photos into natural-sounding audio.
2/19/2026
Githuman
GitHuman tool moves AI code review from pull request to staging area, providing proper review interface before commit for AI-generated code.
2/19/2026
Hankweave: A runtime for data agents, designed for debugging
Hankweave is a single-threaded data agent runtime prioritizing debugging, maintainability, and long-horizon execution over ease of use.
2/19/2026
Security audit of OpenClaw and other similar open source AI Agents
Security audit examining vulnerabilities in OpenClaw and similar open source AI agent frameworks.
2/19/2026
Building an Elite AI Engineering Culture in 2026
Discussion of organizational practices that separate elite AI teams from average ones, based on Faros AI productivity data across 10,000+ developers.