Skip to content
Agent Engineering Lab

Signal

What to keep, what to skip. Commentary on the talk around agent engineering: papers, tools, marketplaces and claims. Field Notes report from inside the work. Signal reads what everyone else is saying.

  1. SG-004 Aug 14, 2026 Loop engineering is regulator engineering The practical manual for running agent loops arrived this week. The theorems for the half that actually decides whether a loop is safe arrived between 1956 and 1970. Nobody in the discourse is citing them. 15 min
  2. SG-003 Jun 7, 2026 Multi-agent is not an architecture decision. It is a workload decision. Anthropic and Cognition were not disagreeing. One measured breadth-first research, the other warned about shared-state work. The variable is decomposability. 10 min
  3. SG-002 May 23, 2026 Long context did not kill RAG. It changed what retrieval is for. Long context made lookup cheap. The harder work, selecting evidence, citing it, governing access to it, and keeping it current, did not move. Here is the routing rule and the literature that supports it. 9 min
  4. SG-001 May 14, 2026 The skills.md discourse is sharpening the wrong knife The community is writing token-saving guides and building million-skill marketplaces. The actual constraint in Anthropic's Skills architecture is a 1% description budget that silently trims the labels of skills you invoke least. 8 min