3/17/2026
Show HN: Driggsby – securely chat with your finances
MCP-based financial data aggregator using Plaid for AI clients. Demonstrates practical MCP application for secure data integration.
3/17/2026
Memory that lives between AI agents, not within them
Memory architecture for AI agents using external shared storage rather than internal models. Title only, minimal detail.
3/17/2026
Remembr – Open-source long-term memory for AI agents
Open-source long-term memory system for AI agents. Title only, lacks implementation details.
3/17/2026
We don't need to hack your AI Agent to hack your AI Agent
Security analysis of AI agent vulnerabilities. Title only, unclear technical depth.
3/17/2026
Agentic Representation of Ecosystems
Conceptual project exploring AI agents representing ecosystem interests and legal rights, combining agent design with environmental protection frameworks.
3/17/2026
Show HN: FC-Eval – CLI to Benchmark Local or Cloud LLMs on Function Calling
CLI benchmark for evaluating LLM function calling across 30 test cases. Supports cloud and local models for agent workflow testing.
3/17/2026
Show HN: Detecting LLM hallucinations in <1ms using hidden states (RTX3050, 4GB)
Open-source tool detecting LLM hallucinations via hidden state analysis. Achieves 0.90+ ROC-AUC on Gemma/Llama with <1ms latency.
3/17/2026
2X-3X productivity from parallel agents
Technique for running multiple parallel AI coding agents simultaneously using git worktrees to achieve 2-3X productivity improvement over sequential execution.
3/17/2026
Show HN: Flock v0.7.0 – Semantic Layer for DuckDB (C++)
Flock v0.7.0: Open-source DuckDB extension enabling LLM operators and RAG pipelines natively in SQL. Adds Anthropic/multi-provider support.
3/17/2026
Show HN: Mengram – Open-source memory layer for AI agents
Open-source memory layer for AI agents. Title only, lacks technical implementation details.
3/17/2026
Unix Reimagined for AI – Iterative Coding in a Shell Loop
Shell-based iterative coding approach for AI. Title only, insufficient detail provided.
3/17/2026
Capyra – open-source agent runtime for SAP B1 and WhatsApp
Open-source autonomous agent runtime connecting AI to business systems (ERP, databases) via WhatsApp, Slack, Telegram with action capabilities.
3/17/2026
Ask HN: With Promptfoo acquired by OpenAI, what are MCP devs using for testing?
Discussion on testing tools for MCP servers after Promptfoo acquisition. MCPSpec project for CI testing of Model Context Protocol.
3/17/2026
Show HN: MUP – Interactive UI inside LLM chat, so anyone can use agentic AI
MUP (Model UI Protocol) enables interactive UI components in LLM chat, allowing both users and agents to trigger functions. Includes PoC host and 9 example implementations.
3/17/2026
The first open-source agentic AI physicist
Open-source AI agent designed to perform physics research tasks autonomously.
3/17/2026
Show HN: Oh-my-agent – A structural harness for AI agents in real projects
Framework for reliable AI agent development addressing hallucination and task drift. Structured protocol for production agent deployments.
3/17/2026
Show HN: MCP Isn't Dead. You're Just Using It Wrong
Analysis of MCP dynamic tool registration feature. Argues MCP enables advanced agent capabilities beyond static tool definitions.
3/17/2026
Ask HN: What AI can you use for personal video editing?
User question about AI tools for personal video editing. Discussion of limitations in current LLM video capabilities.
3/17/2026
LLM drift-Claude vs. Calmkeep: 25-turn Code (60% vs. 85%) & Legal (50% vs. 100%)
Performance comparison of Claude vs Calmkeep on 25-turn code and legal tasks. Shows 60%-85% code accuracy and 50%-100% legal accuracy.
3/17/2026
Three zones of LLM competence for software engineers
Analysis of LLM competence zones for software engineering tasks. Framework for understanding model capabilities and limitations.
3/17/2026
LLMs benchmark with esoteric programming languages
Benchmark study showing LLM code generation relies on memorization. Models score 90% on Python but 3.8% on esoteric languages.
3/17/2026
Show HN: FreeFlow – Open-Source Wispr Flow
Open-source voice-to-text tool with real-time speech cleaning and injection into any app. Customizable alternative to Whisper Flow.
3/17/2026
ClickSay – Chrome extension that captures UI context for AI coding tools
ClickSay is a Chrome extension that captures UI context (selectors, styles, HTML, screenshots) and voice input for AI coding tools like Claude Code.
3/17/2026
Show HN: The new security frontier for LLMs; SIEM evasion
Security research showing AI agents can perform SIEM/EDR evasion, indicating organizations must assume adversaries will gain these LLM-powered capabilities.
3/17/2026
How I Used Lima for an AI Coding Agent Sandbox
Experience report using Lima for sandboxing AI coding agents (Claude Code, Codex) to enable autonomous operation with controlled permissions.
3/17/2026
Introducing GPT-5.4 mini and nano
OpenAI releases GPT-5.4 mini and nano models optimized for coding and subagents with 2x faster inference and improved reasoning.
3/17/2026
Rtk – CLI proxy that reduces LLM token consumption by 60-90%
Rtk is a Rust CLI proxy reducing LLM token consumption 60-90% by filtering and compressing command outputs before context, with <10ms overhead.
3/17/2026
Ask HN: Quick technical questions about LLMs
Discussion thread with technical questions about LLM mechanics: token stopping, prompt continuation, and next-token prediction behavior.
3/17/2026
Can LLMs Hack Enterprise Networks?
Prototype using LLMs for autonomous assumed-breach penetration testing against Active Directory networks, demonstrating LLM capabilities in enterprise security contexts.
3/17/2026
A structural epistemic limit in LLMs: 8–15% unverifiable claims across domains
MarCognity-AI is an open-source framework analyzing LLM claim verification, finding 8-15% unverifiable claims. Decomposes responses and verifies against sources.
3/17/2026
Out-of-Context Reasoning in LLMs: A short primer and reading list
Primer on out-of-context reasoning in LLMs: when models reach conclusions requiring reasoning not present in context window, affecting generalization and alignment.
3/17/2026
Show HN: ModelSweep - Open-Source Benchmarking for Local LLMs
ModelSweep is a GUI-based benchmarking workbench for evaluating local LLMs on Ollama, enabling test suite building and comparative dashboards.
3/17/2026
Llmgate – call any LLM via YAML config, 2 dependencies
Llmgate is a lightweight Python wrapper supporting 21 LLM providers via YAML config with only 2 dependencies (httpx, pyyaml).
3/17/2026
Show HN: DataFlow,Turn raw data into high-quality LLM training datasets
DataFlow is a low-code visual pipeline tool for generating, cleaning, and preparing high-quality LLM training datasets with flexible orchestration.
3/17/2026
Show HN: MCP Inspector – connect and test any MCP server
Web-based MCP server inspector tool for testing Claude MCP implementations with shareable URL configurations.
3/17/2026
RetroAgent: From Solving to Evolving via Retrospective Dual Intrinsic Feedback
Online RL framework enabling LLM agents to evolve through retrospective dual intrinsic feedback and self-reflection. Code released.
3/17/2026
Human Attribution of Causality to AI Across Agency, Misuse, and Misalignment
Study investigating how humans attribute causal responsibility in AI-related harmful incidents involving agency, misuse, and misalignment.
3/17/2026
A Dual-Path Generative Framework for Zero-Day Fraud Detection in Banking Systems
Dual-path generative framework for real-time fraud detection in banking that balances low-latency anomaly detection with GDPR explainability requirements.
3/17/2026
Benchmarking Zero-Shot Reasoning Approaches for Error Detection in Solidity Smart Contracts
Benchmark evaluating zero-shot LLM prompting strategies for detecting security vulnerabilities in Solidity smart contracts.
3/17/2026
Think First, Diffuse Fast: Improving Diffusion Language Model Reasoning via Autoregressive Plan Conditioning
Plan conditioning method improving diffusion language models' multi-step reasoning by prepending autoregressive plans to guide iterative denoising.
3/17/2026
Automating Document Intelligence in Statutory City Planning
AI system automating document intelligence for UK planning authorities to navigate Data Protection Act and Planning Act compliance in large document volumes.
3/17/2026
Multi-Axis Trust Modeling for Interpretable Account Hijacking Detection
Multi-axis interpretable trust framework for detecting account hijacking using multidimensional criteria instead of single anomaly scores.
3/17/2026
ILION: Deterministic Pre-Execution Safety Gates for Agentic AI Systems
Safety framework for agentic AI systems that enforces deterministic pre-execution checks on real-world actions before API calls and transactions execute.
3/17/2026
ManiBench: A Benchmark for Testing Visual-Logic Drift and Syntactic Hallucinations in Manim Code Generation
ManiBench benchmark for evaluating LLM code generation quality for Manim animations, testing syntactic hallucinations and visual-logic correctness.
3/17/2026
Distilling Deep Reinforcement Learning into Interpretable Fuzzy Rules: An Explainable AI Framework
Hierarchical fuzzy rule distillation framework for interpreting deep reinforcement learning policies in safety-critical applications.
3/17/2026
Multi-hop Reasoning and Retrieval in Embedding Space: Leveraging Large Language Models with Knowledge
Method integrating knowledge graph retrieval with LLMs to improve multi-hop reasoning and reduce hallucination through symbolic knowledge enhancement.
3/17/2026
Agent-Based User-Adaptive Filtering for Categorized Harassing Communication
Agent-based framework for personalized harassment filtering in social networks using adaptive filtering agents that learn from user feedback.
3/17/2026
DOVA: Deliberation-First Multi-Agent Orchestration for Autonomous Research Automation
Multi-agent LLM platform with deliberation-first orchestration for autonomous research automation combining tool use, reasoning, and code generation.
3/17/2026
Why Grokking Takes So Long: A First-Principles Theory of Representational Phase Transitions
First-principles theory explaining the delay between memorization and generalization in neural network grokking with weight decay analysis.
3/17/2026
DyACE: Dynamic Algorithm Co-evolution for Online Automated Heuristic Design with Large Language Model
LLM-based automated heuristic design system that dynamically evolves algorithms for non-stationary combinatorial optimization tasks.