Prompt Pack
Prompt Pack

Skip-Level

GitHub: Agents Started More Than 40 Percent of September Pull Requests

By ยท 10 Nov 2026

An IT manager, seen from behind, stands at an ice blue counter and looks at a landscape monitor with a dense grid of blank ivory and persimmon tiles, one ivory tile set slightly apart, against a bright emerald green wall.
  • GitHub's Octoverse 2026 report, updated Nov 6, finds that by September more than 40% of pull requests started with an agent's first draft, up from 17% in March.
  • That totaled 110 million agent started pull requests in September alone, about six times the March volume.
  • An August 14 Gartner guest post on CIO Dive says only 21% of C suite executives view their peers as AI savvy.
  • Anthropic released Claude Fable 5.2 and Mythos 5.2 on Nov 9, with Fable available to all users and Mythos limited to verification programs.

Three things engineering and IT managers should know this Tuesday, from the share of pull requests that now begin as an agent draft to whether your executive peers can govern the work those agents produce.

1. GitHub: Agents Started More Than 40 Percent of September Pull Requests

GitHub published its Octoverse 2026 report on Oct 28 and updated it on Nov 6. The report says developers pushed code to GitHub 17.9 billion times in the past 12 months, more than in 2024 and 2025 combined. The finding that should change how you staff review is this: by September 2026, more than 40% of pull requests on GitHub started with an agent's first draft, up from 17% in March. That is 110 million in September alone, about six times the March volume. GitHub defines an agent started pull request as one whose first draft came from a coding agent it can detect, and it calls that figure a floor because not all agent use may be detected.

The report also says that for a growing share of work, the first move is now directing an agent, and review is how developers shape what comes back. Among repositories created this year, the average repo with an agent configuration file logged eight times as many commits as the average new repo without one. Agent started pull requests more often change multiple files, about 87% compared with 79% of pull requests without detected agent involvement.

Why it matters: GitHub is describing a shift in where the scarce time goes, from writing the first draft to reviewing what an agent produced, at a volume of 110 million agent started pull requests in a single month. Our read: treat review capacity, verification skills, and merge standards as the constraint, not raw coding output. Ask how many of your team's pull requests now begin as an agent draft, who is accountable for the review, and whether your current review load can absorb a further rise. Source: GitHub

2. Gartner Said in August That Few Executives See Their Peers as AI Savvy

This one is older, so the date matters. In an August 14 guest post on CIO Dive, Gartner director analyst Gladys Yeo wrote that executive AI literacy remains underdeveloped even as organizations set ambitious AI goals. Just 21% of C suite executives view their peers as AI savvy, according to Gartner data cited in the post. Yeo describes executive AI literacy as the combination of working knowledge of data and AI and the ability to apply executive judgment to AI driven decisions. She argues that most organizations focus on general employee AI literacy and leave executives with fragmented or generic self directed learning. Gartner research, she writes, shows organizations that prioritize executive AI literacy are better able to convert AI ambition into sustained business value. The post asks CIOs to help executives build that judgment across four areas: governing AI, prioritizing investment, preparing the workforce, and positioning the business for long term advantage.

Why it matters: The 21% figure is Gartner's own data as presented in an August guest post, not a new November survey, and it measures how executives view their peers rather than a direct skills test. Our read: if more than 40% of pull requests already start as an agent draft, the people who set policy and budget need enough working knowledge to govern that work, not only personal use of generative AI tools. A practical step is to brief your own manager on what share of your team's pull requests now start with an agent, what you review before merge, and which decisions you need them to be able to challenge. Source: CIO Dive

3. Fable 5.2 Raises a Planning Question About Who Gets Which Model

Anthropic announced Claude Fable 5.2 and Claude Mythos 5.2 on Nov 9. It says Fable includes cybersecurity and biology safeguards that let the company make it available to all users, while organizations pursuing beneficial applications in cybersecurity or biomedical research can apply for Mythos through Anthropic's Life Sciences and Cyber Verification Programs. Anthropic also says Fable and Mythos are best for the most ambitious, open ended technical work, and that for most coding and knowledge tasks users should start with Claude Opus 5.5, then move up to Fable or Mythos or down to smaller models as the work requires. It lists the 5.2 models at $8 per million input tokens and $40 per million output tokens, which it describes as a lower price than the prior generation, and it says they generate outputs up to 30% faster than the 5.1 models.

Why it matters: The planning issue is not which editor ships the model first. It is who inside your organization is approved to use which model, for what kind of work, and under what review rules. Our read: write down whether Fable 5.2 is even the right default for most of your teams, given Anthropic's own guidance to start with Opus 5.5, and who would have to apply for Mythos if a cybersecurity or life sciences workflow needs it. Revisit spend and approval paths if longer agent jobs become more common, and do not treat the speed and price figures as a reason to skip a pilot. Source: Anthropic

Compiled daily with Grok from public sources and edited by the AIDevPulse team.

Get Skip-Level and the daily brief in your inbox: sign up for the AIDevPulse newsletter. For today's developer news, read today's main brief.

Sources:

Recommended AI prompt

Copy this paragraph into ChatGPT, Claude, Gemini, Grok, or whatever AI you use.

Prompt

I manage an engineering or IT team. Using only the three items in this readout, help me draft a one page plan for the next month that does three things: measures what share of our pull requests now start with an agent's first draft and whether our review capacity can keep up, briefs my manager on the working knowledge and judgment our executives need in order to govern, prioritize, and measure that work, and sets a simple access policy for who may use Fable 5.2, who would need Mythos access through Anthropic's verification programs, and when we should stay with a daily driver model such as Opus 5.5. Keep my team's context in mind: [describe your team's size, the tools you use today, and which models you already allow]. Do not add facts that are not in the readout, and flag anything I should check before sharing.