CommunityHacker News·4d ago·3 min readSummaryListen 0:43
Stop running everything through LLMs: How I fixed my research agentA research agent was improved by reducing unnecessary LLM calls in its pipeline.
News Hub
What actually shipped in agent engineering, pulled from the labs, arXiv and Hacker News.
See who we follow41 articles
Launch HN: Kino AI (YC S23) – AI media manager and video editor
Kino AI is an AI media manager and video editor startup from Y Combinator S23.
CommunityHacker News·7d ago·3 min readSummaryListen 0:53
Open models can do 90 percent of agentic work at a twentieth of the costOpen models perform ninety percent of agentic work at one-twentieth the cost of proprietary models.
Launch HN: Momentic (YC W24) – Agents that catch bugs by using your app
Momentic is a YC W24 agent that catches bugs by using applications.
R/open source bans all solo projects and projects using LLMs
R/open source subreddit banned solo projects and projects using LLMs.
CommunityHacker News·19d ago·3 min readSummaryListen 0:44
LLM agents play Pokemon showdownLLM agents play Pokemon Showdown.
Launch HN: Alter (YC S25) – Access control for multi-user agents
Alter is a YC S25 startup providing access control for multi-user agents.
Launch HN: Respan (YC W24) – Open-source Rust AI gateway built around plugins
Title alone insufficient for concrete summary.
Launch HN: Laminar (YC S24) – We trained a model to catch silent agent failures
Laminar trained a model to detect silent failures in AI agents.
Hybrid Reasoning Agents: LLM agents that think in latent space, act through text
Hybrid Reasoning Agents enable LLM agents to think in latent space and act through text.
Also covered byarXiv cs.AI
CommunityHacker News·27d ago·3 min readSummary
Show HN: Jevcrawl – A web crawler for AI agents with no LLM in the crawl loopJevcrawl is a web crawler for AI agents that does not use an LLM in the crawl loop.
Launch HN: Consus (YC F24) – LLM gateway for companies handling sensitive data
Consus is an LLM gateway designed for companies that handle sensitive data.
Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agents
Magnitude, a Y Combinator S25 startup, launches a self-optimizing inference engine for agents.
Show HN: Open-source model routing for coding agents at Astra-level performance
Open-source model routing for coding agents achieves Astra-level performance.
Open sourceHacker News·42d ago·3 min readSummaryListen 0:50
Show HN: OpenAPPA – open-source deterministic guardrails that don't break agentsOpenAPPA is an open-source project providing deterministic guardrails for AI agents without breaking their functionality.
Understanding the Impact of LLM Watermarking on AI Agent Behavior
Study examines how watermarking techniques added to large language models affect the behavior and performance of AI agents.
Launch HN: Coverage Cat (YC S22) – Umbrella insurance via your personal agent
Coverage Cat uses personal AI agents to sell umbrella insurance.
Show HN: CUA-S1 – A System One Model for Computer Use
CUA-S1 is a System One model designed for computer use tasks.
CommunityHacker News·56d ago·3 min readSummaryListen 0:43
A1ex: A simple LLM coding agent in LuaA1ex is a coding agent built with large language models, implemented in the Lua programming language.
Multi-Agents LLM Financial Trading Framework
Multi-agent LLM framework developed for financial trading applications.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.