3/31/2026
Show HN: The 42-Day Vibe – A mockumentary on "Vibe Coding" taken to the extreme
Mockumentary podcast about LLM agent managing codebase autonomously for 42 days. Explores AI agents in software development.
3/31/2026
Show HN: Multi-agent autoresearch for ANE inference beats Apple's CoreML by 6×
Multi-agent system optimizing Apple Neural Engine inference 6× faster than CoreML through collaborative learning. AI agents research.
3/31/2026
Show HN: Cerno – CAPTCHA that targets LLM reasoning, not human biology
Discussion thread about software developer burnout.
3/31/2026
Encoding Team Standards
Content stub on United Airlines design system.
3/31/2026
Show HN: I adapted codex-plugin-cc's design for Gemini CLI's ACP
Open-source TypeScript SDK for CAPTCHA that targets LLM reasoning capabilities rather than human biology.
3/31/2026
Major Claude Code source leak offers deep insight into how Anthropic tool works
Semantic search tool for 29,657 HuggingFace models with natural language queries, optimized for finding small specialized models without embeddings.
3/31/2026
Show HN: Wozcode – double Claude Code output
Legal analysis of BIPA lawsuits against AI note-taking software vendors over biometric data collection without consent.
3/31/2026
Jailbroken, open-source, pre-built iOS 26 virtual machine
Agent harness tool benchmarked 40% faster and 25-55% cheaper than Claude Code, now available as Claude Code plugin for mobile app building.
3/31/2026
Ask HN: I burnt out from software development. What now?
Platform enabling autonomous AI agents to transact using open protocols with cryptographic verification and reputation scoring.
3/31/2026
LLMs encourages delusional thinking in patients, study finds
Windows application for command-palette style app switching inspired by macOS rcmd.
3/31/2026
Do LLMs Break the Sapir-Whorf Hypothesis?
Research examining whether transformer-based LLMs break the Sapir-Whorf hypothesis about language shaping thought through analysis of reasoning layers.
3/31/2026
GitHub Will Train Copilot Models with User Data by Default
GitHub policy change defaulting to use Copilot interaction data for model training with opt-out available starting April 2026.
3/31/2026
Sprintra – Persistent memory for AI coding agents (MCP-native)
Sprintra adds persistent memory capability to AI coding agents using MCP protocol.
3/31/2026
Claude Code fork that works with any OpenAI-compatible LLM
Open source fork of Claude Code supporting multiple LLM backends (OpenAI, Ollama, vLLM, Groq) with agent toolchain for file editing and task automation.
3/31/2026
Show HN: AI Engineering AI-native Self-learning course repo
Self-learning course repository for AI engineering, appears to be educational content/curriculum.
3/31/2026
CoPaw – A truly local OpenClaw alternative with customized models (2B, 4B, 9B)
CoPaw local AI assistant supporting 2B-9B models with self-hosting, memory control, and skill extensions without vendor lock-in.
3/31/2026
Pipevals: Evaluation pipelines for every LLM application
Pipevals: open-source evaluation pipeline builder for LLM applications. HTTP API integration for quality tracking without SDK wrapper.
3/31/2026
Why Nvidia paid $20B for Groq instead of fixing the GPU
Headline only, no content provided.
3/31/2026
Ask HN: Has anyone benchmarked skill availability impact on LLM performance?
HN discussion asking if anyone systematically benchmarks LLM skills/tools to measure their impact on model performance.
3/31/2026
Add the missing source to Claude Code
Incomplete GitHub PR discussion with error messages, low coherence.
3/31/2026
Get video proof of what your agents did in TUIs
Tool recording AI agent interactions with TUIs as video proof, showing keystrokes and system state with full ANSI color support.
3/31/2026
Show HN: Dograh – voice agents that pick Recordings over TTS using LLM
Dograh open source voice agent platform with drag-and-drop workflows, Gemini 3.1 support, and pre-recorded audio for local deployment.
3/31/2026
'Worst in Show' CES products include AI fridges, AI companions and AI doorbells
News article about poorly-reviewed AI consumer products from CES including fridges, doorbells, and companions.
3/31/2026
Show HN: Domscribe – Deterministic DOM-to-source mapping for AI coding agents
Domscribe tool providing deterministic DOM-to-source mapping for AI coding agents to reduce token usage in frontend automation tasks.
3/31/2026
Open-source security tests for AI agents (MCP, A2A, x402, AIUC-1)
Open source security testing framework for multi-agent AI deployments covering agent escalation, context poisoning, and prompt injection attacks.
3/31/2026
Show HN: An agent skill that tracks SF city hearings, permits, lobbying etc.
OpenClaw skill aggregating 22 SF government data sources to track city hearings, permits, and lobbying decisions.
3/31/2026
Show HN: SuprLogs – Autopilot changelogs from GitHub commits
Tool generating automated changelogs from GitHub commits using LLM analysis of diffs for non-technical audiences.
3/31/2026
A bug in Bun may have been the root cause of the Claude Code source code leak
Report of Bun bundler bug that exposed source maps in production, potentially related to Claude Code source leak.
3/31/2026
Show HN: OpenClaw Arena – Benchmark models on real tasks, rank by perf and cost
OpenClaw Arena benchmark platform comparing AI models on real agentic tasks with tool access, separate leaderboards for performance and cost.
3/31/2026
How we made Trail of Bits AI-native (so far)
Case study on Trail of Bits' AI adoption strategy, building integrated system for 94% employee adoption with 94 plugins.
3/31/2026
Show HN: Open-FDD – On Prem HVAC Fault Detection with OpenClaw, BACnet, & Brick
Open-source on-premises knowledge graph platform for HVAC fault detection using BACnet and Brick, enabling facility energy optimization.
3/31/2026
How HN: Synthetic You – AI calibrated to your personality
Service offering personality-calibrated AI mentor prompts based on psychometric assessments for ChatGPT and Claude.
3/31/2026
AI agent is authorized to do everything wrong
Technical article on authorization failures for autonomous AI agents in production systems, using invoice processing pipeline example.
3/31/2026
Show HN: Shards of Stone – A Warcraft 2-inspired RTS that runs in the browser
Browser-based RTS game inspired by Warcraft 2 with multiplayer and AI opponents, built using Claude as development tool.
3/31/2026
Show HN: Fakekey – never expose real API keys in the agent environment
Fakekey open source project preventing API key exposure in AI agent environments through token masking and secure credential handling.
3/31/2026
Show HN: Dewey – Ingest docs, search semantically, get cited AI answers
Dewey tool for ingesting documents and semantic search with hierarchical chunking for RAG and multi-hop reasoning across documents.
3/31/2026
Show HN: LLminate – LLM Detector
Proof-of-concept LLM detector leveraging training data properties to identify LLM-generated content.
3/31/2026
Tell HN: Blogburst.ai sent me AI spam message email to my Show HN project
Complaint post about receiving unsolicited AI-generated spam email after publishing a Show HN project.
3/31/2026
Show HN: Gha-outrunner – Ephemeral GitHub Actions runners without Kubernetes
Go tool providing ephemeral GitHub Actions runners using Docker/KVM/Tart without Kubernetes, open source MIT licensed.
3/31/2026
Show HN: I built a visual ML trainer that exports to standalone PyTorch code
MLForge visual ML trainer exports to standalone PyTorch code. Node-based interface with GPU/CPU/MPS support and early stopping.
3/31/2026
Multimodal file search and sharing for AI agents
Title-only entry on multimodal file search and sharing for AI agents. Insufficient content.
3/31/2026
Estimating ISS speed from images using OpenCV (~2–3% error)
Python project using OpenCV to estimate ISS speed from images with ~2-3% error. Feature detection and matching approach.
3/31/2026
Krazam – Offsite Karaoke [video]
Karaoke video. Off-topic.
3/31/2026
Strengthening GitLab.com security: Mandatory multi-factor authentication
GitLab implementing mandatory multi-factor authentication for user accounts and API endpoints as security measure.
3/31/2026
Show HN: SkillForge – Turn code and docs into instructions AI agents can follow
SkillForge converts code and documentation into instructions for AI agents to execute.
3/31/2026
Using LLMs to amplify human labeling and improve Dash search relevance
Dropbox Dash uses LLMs with RAG pattern for enterprise search. LLMs amplify human labeling to improve retrieval and response relevance.
3/31/2026
Claude Code Internals: An AI-Assisted Analysis of the Leaked Source
Title-only entry analyzing Claude Code internals from leaked source. Insufficient content.
3/31/2026
Show HN: TraceHouse – ClickHouse Monitoring
Open-source ClickHouse monitoring tool with visual exploration of queries, merges, resource attribution, and cluster workload in real-time.
3/31/2026
Show HN: Agent Wellbeing Kit – boundary protection for humans running AI agents
Title-only entry on safeguards for humans operating AI agents. Insufficient content.
3/31/2026
Autodebug: Telemetry-Driven Inference Optimization Loop
Autonomous agent continuously deploying inference services, benchmarking, analyzing telemetry, and redeploying with optimized configurations.