Fetching from the wire…
01
02
03
04
05
06
07
08
Vibe Coding2026-08-29 · source-backed
The method turns tool use from a hardcoded prompt into a learned runtime behavior, then applies cost-aware RL teaching the agent when reading external state is worth the token budget. Qwen3-8B reaches a 96.9% average success rate against SkillOS at 80.2% and SkillRL at 89.9%, within two points of Claude Opus 4.5, and the same harness pushes Opus 4.5 to 98.5%. The claim builders should test rather than accept: harness design, not model size, is where the remaining points live. (VentureBeat)
Each link below shares sources, entities, or timing with this story.
Claude Code uses Opus / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code uses Opus); both cover Claude Opus, Meta, Opus; overlapping topics (agent, claude, harness, opus).
Meta partners with Google / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Meta partners with Google); both cover Claude Opus, Meta, Qwen3; overlapping topics (average, claude).
Meta partners with Google / Shared entities / Same source domain / Earlier coverage
Linked by a graph relationship (Meta partners with Google); both cover Claude Opus, Meta, VentureBeat; reported by the same outlet (venturebeat.com).
Meta released Muse Glimmer / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Meta released Muse Glimmer); both cover Meta, Opus; overlapping topics (agent, claude, harness, point).
Qwen competes with Meta / Shared entities / Earlier coverage / Tension
Linked by a graph relationship (Qwen competes with Meta); both cover Claude Opus, Opus, Qwen3; earlier Claude Opus coverage from 2026-04-23.
Linked by a graph relationship (Qwen competes with Meta); both cover Claude Opus, Opus, Qwen3; earlier Claude Opus coverage from 2026-04-21.
Claude Code uses Opus / Shared entities / Same source domain / Earlier coverage / Tension
Linked by a graph relationship (Claude Code uses Opus); both cover Opus, VentureBeat; reported by the same outlet (venturebeat.com).
Claude Code uses Opus / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code uses Opus); both cover Opus, Qwen3; overlapping topics (agent, claude, opus).