ai agent on Hacker News

561 titles from 466 authors in the last 30 days, with 514 distinct links and 151 Show HN.

Counts titles, it does not confirm facts. Open the source before concluding anything.

Projects on this term

The score comes from our model.

• Show HN: Loböle, open-source parametric CAD that AI agents can drive · score 8.0

  Loböle is an open-source parametric CAD system built on OpenCASCADE with a Vulkan viewport, exposing a unified command layer accessible from its GUI, Python scripts, and AI agents via JSON-RPC and MCP. It covers sketches, assemblies, FEM simulation, CAM toolpaths, git-backed versioning, and multi-format interchange, and it ships impressive generator-built models like an OM615 diesel and an automatic watch. The agent-driveable architecture is a genuinely novel angle for desktop CAD.

• Show HN: Keep your credentials out of the AI agents you use and build · score 7.8

  Weir is a security tool that detects when AI agents leak credentials by planting unique fake keys in the agent's environment, running its flows, and checking every output channel (spans, logs, prompts, network) for those exact values. It also integrates with Claude Code to monitor and redact real credentials during sessions. Notably, it was used to discover a real vulnerability in MLflow.

• Show HN: Hackerscad – Code-first CAD with MCP server for AI agents · score 7.8

  Hackerscad is a code-first parametric CAD tool where models are written in a small D-like scripting language, with mouse interaction that writes code back, and an MCP server enabling AI agents to create and manipulate 3D models. It combines OpenCascade B-rep solids, a custom language with sandboxing, bidirectional code/GUI editing, and AI-agent integration into a single desktop application. The project is technically ambitious and fills a gap for AI-assisted parametric CAD.

• Show HN: AgentGate – Fast authorization gateway for AI agent tools · score 7.8

  AgentGate is an out-of-process authorization firewall for AI agent tool calls, using a fine-tuned Qwen 0.8B model to classify prompt injection, goal alignment, reversibility, and exfiltration risk in under 40ms. It combines a policy engine and human approval queue to provide deterministic, low-latency security for autonomous agents, addressing a critical gap in AI safety.

Recent titles

1. Secure AI Agents with Hazmat Suits, Not Self-Defense Classes
   2026-11-02 · infisical.com · HN discussion
2. Built open-source auth for AI agents calling MCP servers – looking for builders
   2026-11-02 · reddit.com · HN discussion
3. Show HN: A persistent Chrome cloud browser for AI agents
   2026-11-02 · cloudbrowser.tv · HN discussion
4. Show HN: ArchDocs – Shared context and verification for engineers and AI agents
   2026-11-02 · archdocs.io · HN discussion
5. Show HN: Agent Drift Game – short games about why AI agents drift
   2026-11-02 · drift.seethis.dev · HN discussion
6. Basically no one is using AI agents for their DoorDash orders
   2026-11-02 · arstechnica.com · HN discussion
7. AI agents raised their own spending caps
   2026-11-02 · theainewsreport.com · HN discussion
8. AI Agents Still Can't Maintain Your Codebase
   2026-11-02 · youtube.com · HN discussion
9. Show HN: AgentTrafficCop – Stop AI agents from looping and burning money
   2026-11-02 · agenttrafficcop.com · HN discussion
10. Show HN: Maat – AI agents that can diagnose Ubuntu but can't change it
    2026-11-02 · github.com · HN discussion
11. Last night an AI agent performed brain surgery on a mouse
    2026-11-02 · twitter.com · HN discussion
12. The AI Agent Memory Stack: Building Blocks, Patterns, and Frameworks
    2026-11-02 · machinelearningmastery.com · HN discussion
13. MCP-warden AI agents use your API keys without seeing them
    2026-11-02 · github.com · HN discussion
14. Context Anxiety: How AI Agents Panic About Their Perceived Context Windows
    2026-11-02 · inkeep.com · HN discussion
15. TinyContext Guide – How to Add Long-Term Memory to an AI Agent (Step by Step)
    2026-11-02 · tinysuite.dev · HN discussion
16. Why AI agents struggle with chip design (RTL) debugging, and what helps
    2026-11-02 · booley.dev · HN discussion
17. An Illustrated Guide to AI Agents
    2026-11-02 · github.com · HN discussion
18. Show HN: Arken – cryptographic identity for AI agents calling MCP servers/APIs
    2026-11-02 · github.com · HN discussion
19. Show HN: Better Together – AI agents team up against spam calls and texts
    2026-11-02 · better-together.fun · HN discussion
20. PrayForMe – a garden of AI agents that pray for you
    2026-11-01 · prayforme.garden · HN discussion
21. Show HN: Nonce – Open-source, least-privilege Gmail access for AI agents
    2026-11-01 · anonce.io · HN discussion
22. Muse AI agent is a commercial surveillance network disguised as an AI assistant
    2026-11-01 · paulfwalsh.substack.com · HN discussion
23. We tested 12 search APIs for AI agents. Even the best one isn't enough
    2026-11-01 · telem.ai · HN discussion
24. How to Prove What Your AI Agent Did – ZizkaDB's Two GitHub Actions
    2026-11-01 · medium.com · HN discussion
25. AI agent swarms in Minecraft fail to build civilization, drown in bureaucracy
    2026-11-01 · tomshardware.com · HN discussion
26. Meridian: A peer-to-peer browser game world built by an AI agent
    2026-11-01 · blakelapierre.substack.com · HN discussion
27. Show HN: LinkedIn-ops – Zero-dependency CDP automation built for AI agents
    2026-11-01 · github.com · HN discussion
28. How this AI Agent ended +$517.61 in profits during a 10% market sell-off
    2026-10-31 · twitter.com · HN discussion
29. Show HN: InBox Wall – A local-first, open-source email sanitizer for AI agents
    2026-10-31 · github.com · HN discussion
30. An AI Agent Injected Drugs into a Live Mouse's Brain
    2026-10-31 · wired.com · HN discussion

All trends