vibehacker
News
arXiv (KAIST) ·

KAIST's RunningTab keeps a task ledger outside the agent's context so read facts stop going missing

The environment, not the model, logs every file a terminal agent reads, tracks listed-but-unopened files, and blocks the first finish while requirements are still open; the authors found about one in five values an agent had already seen never made it into the deliverable. With the same Pi-based harness it beat plain direct-workspace agents and model-kept TODO lists on Workspace-Bench, TheAgentCompany, and OfficeQA Pro, lifting Gemini 3.8 Flash's Workspace-Bench pass rate from 31.4% to 46.9%.

Read at arxiv.org

More news

View all

Sakana AI's MASS has one model design, run, and judge its own agent teams, then train on the best runs

A single shared model plays executor, workflow optimizer, and evaluator, searches for better multi agent workflows, and fine tunes on executions of the team it judged best. After two cycles with Qwen3.6 27B, score per output token reached 1.2 to 1.6x the base model on four research benchmarks, and the authors found the optimizer, not the evaluator, was the bottleneck; paper and code are open…

Sakana AI

Claude Code Mode plugin lets Claude write one sandboxed JS program to chain your MCP tools

Instead of calling MCP tools one by one and dumping every full result into context, the model writes a single JavaScript program that calls the tools, filters and joins the output in an isolated Node process, and returns only the final answer. It needs Node.js 22.13+, keeps normal MCP permission checks, and installs with /plugin marketplace add gabe4coding/claude code mode…

GitHub / gabe4coding (via Show HN)

Karpathy: the gap in how people understand LLM capability is now a sharp funnel, and it's widening

By his napkin math, about 6B people have barely touched LLMs, 1 to 2B use free tier chatbots, roughly 20M see agents finish weeks long math and code projects from a prompt, and about 5,000 inside labs see swarms of agents working for weeks. He blames scale of ambition, capability peaking in verifiable domains like code, and access tiers, and warns that human review is starting to fall behind…

Andrej Karpathy on X (via Hacker News)

Spotted something we missed? Start a thread.