We run scheduled news sweeps and publish only the items that clear the editorial threshold.
Razenshteyn interviewed what he calls a couple of dozen engineers and researchers, from frontier labs to finance to small start-ups, about how they use coding agents at work. He dropped a planned survey because the stories were too…
An Apache 2.0 framework for training a "harness adapter", an external layer that adapts an existing agent without changing its model or harness. It lists GEPA, SkillOpt, MetaWorld Model, Harness² and Continual Harness as supported methods.
GitHub reports that by September 2026 more than 40% of pull requests started with an agent's first draft, up from 17% in March. That is 110 million in September alone, about six times March's volume.
Snowflake built a 65-task benchmark from its own merged pull requests, graded by the tests that shipped with each change and run on its own Bazel build farm. With opus-5, four harnesses passed 84.1% to 85.1% of trials, a spread the…
A Copilot plugin of skills for setting up feedback loops. /lever loop returns a "Loop Card" listing the metric, how to measure it, how noisy it is and what counts as a win, and an autonomy setting runs from L0, where the agent only…
A two-week paid cohort starting November 9, listed at $1,295. The sales page argues teams should build brakes (automated checks, automated and human review) before accelerators (schedules and triggers that start agents unattended), saying…
Google says Carbon will launch at the same introductory price as Argon, $2 per million input tokens and $10 per million output tokens, with cached input at 95% off. For now it goes to "a subset of cyber defenders" and then to more trusted…
A Debian developer says uploads to the project's NEW queue now appear to be passed to an LLM review without an announcement, quotes a review that calls itself machine-generated and unchecked by the DFSG team, and says he will not upload…
This follows the team's earlier post saying three engineers and a crew of agents merged 1,000 pull requests in one week; we did not read that post. The authors say they tried reviewing every change by hand in early March and abandoned it in…
Cursor says the SDK is the framework behind its own managed agents Bugbot, Security Reviewer and Rollouts, and that it is generally available. An agent is a project directory: prompts in Markdown, tools as TypeScript functions, and evals,…
Leichsenring builds Agent Smith, an open source tool that turns tickets into pull requests, with Claude Code and a rule that he writes none of the code. His report, with figures he counted from the repository on 28 October 2026, lists 1,332…
The 314-page review says it synthesizes 164 scholarly works, 100 practitioner records, 29 benchmark records and 17 records from systems the author operates, and catalogs 206 reliability records. It states its central claim in one line:…
Kreinin, who says he has worked extensively on hardware, argues that unit tests rarely find hard bugs, which tend to come from component interactions and need many input combinations. He says their real job is to stop others from changing…
Ioannou runs an internal company agent on OpenCode 1.18.31 (his deployment, not a benchmark). Token use was climbing faster than sessions or automations could explain.
Ioannou pulled his GitHub contribution counts, private work included, for every month from January 2019. As Principal Engineer he averaged 114 contributions a month from August 2022 to July 2025, against 173 as a full-stack engineer.
Borehole, a code-diligence service, reads each project's commit history for commits that name an AI agent as co-author, and reads its contributing guide, pull-request template and agent instructions for a stated policy. Of the projects with…
The filing states that "over 90% of the contributions to our codebase in September 2026 were written by Claude." It adds that average daily code contributed per engineer in the three months ended September 30, 2026 was over 45 times the…
Scarlett built a factory with five workers: a planner, a builder, a tutorial writer, a reviewer and a previewer. It runs on a persistent remote machine (a Sprite) and, in her words, it is a "transparent, well-lit factory" set against…
The authors read the source of eleven coding harnesses, including Claude Code, Codex CLI, Gemini CLI, OpenHands, Aider, Hermes and OpenCode, plus Databricks' Omnigent as a contrast case. They describe an 83-page paper with 13…
The paper splits an agent into parts that share one durable, append-only log. A driver proposes an action, voters approve or block it, a decider counts the votes, and an executor runs the approved code.