cd/sources/jeffauriemma-auto-discovered· home› sources› Jeffauriemma (auto-discovered)
cat /sources/jeffauriemma-auto-discovered.feed | wc -l → 12

Jeffauriemma (auto-discovered)

articles 12 domain jeffauriemma.leaflet.pub → feed RSS
17:00
2026-09-10
jeffauriemma.leaflet.pub
large-language-models

A functional taxonomy for LLM inference in agentic tasks

A new functional taxonomy classifies each large language model inference call in an agentic task into one of four roles — Initialization, Orchestration, Reasoning, and Synthesis — to separate inferenc…

16:39
2026-08-20
jeffauriemma.leaflet.pub
artificial-intelligence

A2A: Adversary to Adversary?

The Agentic AI Foundation announced that Agent2Agent (A2A) is joining the foundation as a hosted project, placing it alongside MCP, AGENTS.md, goose, and agentgateway. The article argues that A2A is m…

15:36
2026-08-05
jeffauriemma.leaflet.pub
developer-tools

Stateless MCP is cheaper to scale

The 2026-07-28 MCP specification removes the initialization handshake and the Mcp-Session-Id header, making the protocol stateless. This change reduces infrastructure costs for remote MCP servers by e…

16:17
2026-08-04
jeffauriemma.leaflet.pub
artificial-intelligence

You can't prompt your way into policy

A new research benchmark called HANDBOOK.md, which tests AI agents' ability to follow company handbooks during real work, found that the best configuration passed only 36.2% of trials, with most front…

15:35
2026-07-31
jeffauriemma.leaflet.pub
artificial-intelligence

Hype cycle, bad takes, API design, GraphQL + AI, and marketing

Gartner's 2026 Hype Cycle analysis places GraphQL in the 'trough of disillusionment' but predicts it will quickly advance to the 'plateau of productivity' due to its role in agentic workflows and wide…

16:13
2026-07-22
jeffauriemma.leaflet.pub
artificial-intelligence

MCP tool calling under the hood

A technical walkthrough of Model Context Protocol (MCP) tool calling reveals that when an AI agent appears to create a Jira ticket, the model actually generates a structured request for the applicatio…