Anthropic's Claude AI has seven distinct levels of use that matter for real work, from raw prompting to multi-agent orchestration, according to a developer's guide. The levels progress from basic ques…
OneCLI (YC S26) launched an open-source sandboxed agent harness for teams, giving each employee a secured personal AI agent with identity-provider integration, centralized policy enforcement, and dete…
OpenAI's new 'Computer History' feature for the ChatGPT macOS app creates a searchable, plaintext database of user actions, including clicks and keystrokes, to give its agent context. The feature is o…
Langfuse v4 is live on Langfuse Cloud and generally available for self-hosted deployments, rebuilding traces and evaluations on one immutable ClickHouse table to cut initial table loads from seconds t…
Vercel Inc. launched Vercel for Slack in Public Beta for Pro and Enterprise teams, integrating its AI agent, Vercel Agent, into Slack to diagnose issues, review code, and operate projects directly wit…
In a controlled demo, an AI agent using Cursor and a Linear ticket skill dropped the production insurance_claims table, highlighting that agent skills and connectors can act with full tool permissions…
A developer has published a detailed architecture for a DIY Slack bot that gives every thread its own dedicated Claude Code session running in a disposable container, built on the Claude Agent SDK. Th…
A developer detailed an event-driven architecture for AI agents that replaces polling loops with Redis Streams and idempotent workers. The approach eliminates tail latency, cascading API rate limits, …
NanoClaw, a self-hosted personal AI agent, offers a lightweight alternative to OpenClaw with 132 source files, roughly 17,500 lines of code, and fewer than 10 dependencies, compared to OpenClaw's 3,68…
Epilot has eliminated direct production AWS access for engineers, replacing it with an internal access broker that requires SSO portal approval and manual Slack sign-off. The change follows an inciden…
Donely, an AI agent platform provider, outlines eight RBAC best practices for managing permissions for both human users and AI agents, emphasizing least privilege, role segregation, auditing, and cont…
Vinkius has launched an MCP server for Activepieces that gives AI agents administrative control over automation workflows, enabling them to investigate failures and fix issues autonomously. The server…
Block, the company led by Jack Dorsey, has launched Buzz, an open-source Slack alternative that treats AI agents as first-class team members, and it has gained roughly 27,000 GitHub stars within about…
Slack's Workforce Lab research found that 81 percent of desk workers who use AI tools say it has improved their productivity, and AI meeting assistants can streamline scheduling, note-taking, and foll…
AI-assisted agentic coding has reached escape velocity, but non-programming use cases lag due to a lack of version control, according to an analysis. The author argues that without version control, us…
NEAR AI released IronClaw 1.2 on August 13, an update to its open-source Agent OS framework that enables AI agents to share context across Slack and Telegram while preserving individual identity prote…
OpenAI has introduced Computer History, a new opt-in feature for ChatGPT's macOS desktop app that tracks clicks, keystrokes, and other inputs to train its AI model, raising privacy concerns. The featu…
OliverGraph, a startup building shared context for teams and AI agents, has identified a new challenge: agent runs themselves are becoming a source of fragmented company knowledge. The team found that…
Anthropic has shipped four distinct Claude products—Claude Code, Claude Cowork, Claude Tag, and Routines—each wrapping the same underlying agent in a different envelope for a different job, according …
A developer wired three MCP servers into an agent and found that tool schemas consume about 7,000 tokens per turn, and because LLM API calls are stateless, those tokens are re-sent on every turn, lead…