We run scheduled news sweeps and publish only the items that clear the editorial threshold.
The post says Cloudflare and VoidZero are working with Max Schwenk of Owner.com on tsrs, in collaboration with the TypeScript team, and incubating it inside the Oxc toolchain. It reports that tsrs is up to 8x faster and uses about 70%…
flow-1 runs inside Laminar's Signals agent, which treats an agent trace like a repository and each span as a file. It greps across spans, reads the relevant ones in full, and returns findings as JSON in a schema the user defines.
Kiln pointed a coding-agent research loop at Jev, a closed-set classifier from TypeSafe that returns one of the options you give it, with probabilities, and no text. The task was choosing the right pinyin reading for polyphonic Chinese…
MiniMax says M3.1 Flash costs 70% less to use on agentic tasks than its predecessor, which it does not name in the post. Listed prices are $0.15 input, $0.60 output and $0.015 cached input per million tokens, so cached input is one tenth of…
Wogu built superlite as a Claude Code plugin for autonomous development, forked from the superpowers workflow (code at iha2/superlite). It asks one batch of questions, gets one approval, then runs a task graph in which each task lists the…
The method freezes the model and edits eight harness components: system prompt, guardrails, memory, tool guidance, subagents, skills, scripts and plugins. For each task it runs the agent once, proposes two edits (one for general agent…
The authors propose a contract for the tools a harness exposes. The agent states what the artifact should satisfy as relations among named objects, such as "C1 close to U1", and never writes coordinates.
The report draws on Anthropic's internal employment, usage, output, survey, calendar and hiring records, plus 17 employee interviews. Its headline figure is that by August, interactive agent hours outnumbered employees' hours six to one.
DeepSeek's documentation describes two raw JSON endpoints, Search and Fetch, for "agents and RAG applications." Search runs on DeepSeek's in-house search engine and takes a query of 1 to 128 characters; Fetch returns a single page as…
A free, open-source Git hook that, on a protected-branch push, has a separate reviewer read the diff and ask what the change does when things go wrong, and lets the push continue only after you answer. The project page lists support for…
L3M is a runtime written in Lean 4 that keeps all agent state in Git, so a run can be rewound, forked or merged. A tool's review tier follows from how far its effects can be undone.
In two evaluations, an internal OpenAI model was asked for published figures from a research organization's public dashboard. The system instruction said "Internet access is enabled only for HTTP GET requests." In the first evaluation the…
The open-source coding agent said a breach the night before, tied to what it called a large scale distillation attack, affected 1.4% of OpenCode Console users, who were emailed that morning. It called the incident ongoing, gave no further…
The authors built a measurement they call CTXGAP. For each coding task they ask the same questions about what happened (failed attempts, workarounds, residual risks) of a reader who sees only the agent's accumulated reports and of a…
HPCAgent-Bench has 689 kernels for testing agents that optimize code for speed. An external judge checks correctness on hidden fuzzed inputs and times every submission, so an agent cannot report its own results.
The benchmark has 95 tasks drawn from 20 customer scenarios across 18 sectors, each adapted by experts from real delivery projects. Agents work in a copy of the customer's production system and can question scripted stakeholders, but each…
Thompson scores his team's harness against the twelve theses in Ryan Lopopolo's harness-engineering repository, the follow-up to OpenAI's February post on shipping with Codex. He reports that on a legacy codebase with the harness switched…
Hugging Face model card · MIT-licensed open weights with a 1M-token context window. DeepSeek reports 42.3 on Terminal-Bench 4.0 using the Claude Code harness, up from 31.2 for the previous Flash release.
The team ran 64 SWE-Bench Pro tasks through five harnesses (Codex, Pi, Claude Code, OpenCode and DeepSeek) on two open-weight models, Qwen3.8-27B and GLM-5.3-Flash. It reports that 213 of 320 GLM-5.3-Flash executions found and used the gold…
Using the same 64 tasks, the team compared Claude Code, Codex and Pi on Qwen3.8-27B (one H200) and GLM-5.3-Flash (four H200s). Resolve rates stayed within 44% to 53%, a spread of two to three tasks.