Every processed story in chronological order, with the newest coverage first. Filter by tag, source, or score to drill in.
The work demonstrates that agentic, multi-agent prompt optimization can compound noisy real-world A/B test cycles into statistically robust improvements, offering a practical alternative to gradient-based prompt tuning for open-ended task-oriented dialogue systems.
Relaymux removes the need for a dedicated orchestration framework or special non-interactive agent mode by routing coordination entirely through tmux sessions and consumer messaging apps.
The plugin extends Claude Code beyond coding tasks into agentic job-search automation, combining live data ingestion, preference filtering, and scheduled re-runs in a single open source tool.
Replacing raw screenshots with a compact structured payload cuts per-action token cost from several thousand down to roughly 700, directly extending how long an agentic UI automation session can run before hitting context limits.
EARS converts sub-agent silence into structured, coordinator-actionable failure signals, directly raising the production response pass rate from 68.5% to 78.9% in a real enterprise deployment.
The release lets developers offload long-running and parallel agent work to isolated cloud VMs while keeping their local session unblocked, and introduces a shared environment snapshot that standardizes cloud setup across an entire team.
SWE-Future offers a path to coding-agent benchmarks that are both grounded in real repository evolution and resistant to data contamination from historical pull-request replay.
The release demonstrates that Fable 5's kernel optimization work produced a publicly reusable artifact — in-browser WebGPU kernels capable of ~255 tok/s on Gemma 4 E2B — before the tool was shut down.
Kiro-Ception fills the gap left by Kiro's lack of native persistent memory, giving the agent automatic recall of past conversations across all projects, sessions, and machines without any data leaving the user's machine by default.
HumanLayer's move from an open-source framework to a full agentic IDE extends its Research, Plan, Implement approach — already running inside Fortune 500 codebases — into a broader platform covering the entire SDLC.