Tag: context-engineering
All the articles with the tag "context-engineering".
-
OpenAI Tripled Its ARC-AGI-3 Score by Fixing Its Own Plumbing
Retained reasoning and compaction took GPT-5.6 Sol from 13.3% to 38.3% on the ARC-AGI-3 public set with 6x fewer output tokens. The engineering lesson is solid. The 38.3% is not comparable to the 30.2% Opus 5 posted five days earlier.
-
Anthropic Deleted 80% of Claude Code's System Prompt. Your CLAUDE.md Is Next.
Two guides published alongside Claude Opus 5 say most accumulated prompt-engineering folklore is now an anti-pattern: delete your verification steps, delete your examples, delete your rules. Here's what survives, and what the deletions cost you.
-
Ralph Is Five Lines of Bash. All the Bugs Are in the Improvements.
Geoffrey Huntley's Ralph Wiggum technique is a while loop that pipes a prompt file into a coding agent. A year later, Anthropic's official plugin version has been renamed by legal and patched twice for bugs that only exist because it keeps state the bash loop never had.
-
Agent Harnesses: A Standard for a Word Nobody Agrees On
Daniel Warfield published an open spec for agent harnesses this month: HARNESS.md, all-caps routing files, and leaf detectors. The format is sensible. The name is a fight it may not win, because most of the industry already uses 'harness' to mean the thing running the agent.