Fetching from the wire…
01
02
03
04
05
06
07
08
Agents2026-08-09 · source-backed
AMAP-ML released it August 4 (463 stars, v0.1.3 on August 7) on a strict Manager/Executor/Auditor split where, in its own words, "only results that pass independent verification enter persistent task state" (GitHub). WeaveBench 51.8% → 80.7% completion. OSWorld 2.0 2.8% → 8.3% full completion. Terminal-Bench 2.1 69.7% → 77.2% using 24% fewer tokens. It wraps Codex and Claude Code via an AgentAdapter protocol, defaults to 30 rounds, and persists goals, verified progress, audit reports, and role trajectories in isolated run directories. Same architectural lesson as HTTP Terminator: the auditor role, not a bigger context window, is what stops multi-hour runs from drifting.
Each link below shares sources, entities, or timing with this story.
Codex competes with Claude Code / Shared entities / Same source / Shared topic / Earlier coverage
Linked by a graph relationship (Codex competes with Claude Code); both cover AgentAdapter, AMAP, Bench, Claude Code; cite the same source (GitHub).
Codex competes with Claude Code / Shared entities / Same source domain / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Codex competes with Claude Code); both cover Bench, Claude Code, Same, Terminal; reported by the same outlet (github.com).
Linked by a graph relationship (Codex competes with Claude Code); both cover Claude Code, Codex, GitHub; reported by the same outlet (github.com).
Claude benchmarked against Codex / Shared entities / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Claude benchmarked against Codex); both cover Bench, Claude Code, Codex, Terminal; overlapping topics (claude, code, codex).
Codex released Windows / Shared entities / Same source domain / Shared topic / Earlier coverage
Linked by a graph relationship (Codex released Windows); both cover August, Bench, Claude Code, GitHub; reported by the same outlet (github.com).
Codex competes with Claude Code / Shared entities / Same source domain / Shared topic / Earlier coverage
Linked by a graph relationship (Codex competes with Claude Code); both cover Claude Code, Codex, GitHub, Same; reported by the same outlet (github.com).
Linked by a graph relationship (Codex competes with Claude Code); both cover Claude Code, GitHub, Harness, Same; reported by the same outlet (github.com).
Codex competes with Claude Code / Shared entities / Same source domain
Linked by a graph relationship (Codex competes with Claude Code); both cover August, Bench, Claude Code, Codex; reported by the same outlet (github.com).