Electron: The Invisible Infrastructure Behind ChatGPT, Claude, and VS Code

1、First, it is important to define the subject accurately. Electron is not a “Mac-only framework”; it is a cross-platform desktop application framework for macOS, Windows, and Linux. It bundles Chromium and Node.js into the application binary so that developers can build desktop software with JavaScript, HTML, and CSS. The official repository states that it is used by Visual Studio Code and many other applications. As of June 2, 2026, the main repository has about 122k stars, 17.2k forks, more than 30,000 commits, and its latest stable release is v42.3.0, published on May 26, 2026.

2、On the question of the “founder,” the safest formulation is not “one lone founder,” but rather “one dominant original author plus a project that was team-structured from the beginning inside GitHub.” In public materials, the most clearly identified original author and early technical lead is Cheng Zhao, also known as zcbenz. Kevin Sawicki’s 2015 talk described the project as being “led by Cheng Zhao,” and Cheng’s own website lists Electron under “Created,” saying he “Created the Electron project to power Atom editor.” But if the topic is launch, naming, documentation, community-building, and productization, Kevin Sawicki, Zeke Sikelianos, and the broader GitHub Atom team also matter and cannot be erased from the story.

3、What makes Electron important is not the cliché that it “wraps web apps as desktop apps.” What really matters is that it turned the full engineering reality of cross-platform desktop software—updates, installers, signing, native APIs, debugging, testing, release cadence, security patches, and governance—into a maintainable infrastructure stack. In current official documentation, Electron is described as enterprise-grade and mature, and the official docs explicitly list Slack, Discord, Signal, OpenAI’s ChatGPT, Anthropic’s Claude, Visual Studio Code, Loom, Canva, Notion, and Docker as real-world examples. The docs even claim that most desktop computers on the planet run at least one Electron app. That means Electron is no longer a niche experiment; it is part of real desktop software infrastructure.

4、The limits of public information also need to be made explicit. For Cheng Zhao’s date of birth, birth place, parents, family class background, childhood resources, schools attended, degree completion, personal wealth, and equity holdings, public information is limited and cannot currently be confirmed. As for how much original work he contributed during the NW.js phase, and whether Electron should be described as “rewritten from scratch” versus “partly inherited or derived,” public accounts are inconsistent. The research below distinguishes high-confidence facts from disputed claims.

1、Electron’s “family background,” in project terms, is its origin inside Atom. Kevin Sawicki’s history talk places the project start in January 2013, while Electron’s official ten-year retrospective dates the first commit in the electron/electron repository to March 13, 2013. It was originally called Atom Shell, not Electron, and it was built by GitHub as the runtime underpinning for the Atom editor. Atom entered public beta in April 2014. On April 23, 2015, Atom Shell was officially renamed Electron. That renaming mattered because it marked a shift from being an internal Atom technology to being a consciously independent public framework for outside developers.

2、Cheng Zhao’s personal background is sparsely documented in public, but his professional trajectory is unusually clear. His current personal website says plainly, “My name is Cheng and I work at Apple on machine learning.” The site lists Electron, Yue, Chie, node-mlx, and Compilets under “Created,” and lists Atom and NW.js under “Worked on.” It also states that he “Created the Electron project to power Atom editor” and describes his NW.js work as having “Rewrote the project from scratch.” This strongly suggests that his long-term interest has not been a single product, but the problem of bridging JavaScript with low-level systems, GUI runtimes, and native capabilities. That is an inference, but it is a well-supported one because the pattern runs continuously from NW.js, Electron, Yue, and Yode to node-mlx and Apple’s MLX.

3、Education and family details are mostly unavailable, but some early career milestones can be reconstructed. In a 2016 public statement on the history of node-webkit, Roger Wang wrote that he interviewed and recruited “Zhao Cheng” as an intern in 2012 to work on node-webkit; the same statement says that GitHub later contacted the team and that Cheng then left for Atom Shell. Smashing Magazine’s 2017 article on desktop web frameworks also described the early phase as one in which “an intern named Cheng Zhao joined the project.” What can be confirmed is that he became involved very early in the integration of Node.js and Chromium. What cannot be confirmed is the earlier family and school background.

4、If one asks for Cheng Zhao’s first representative professional experience, the answer is not Apple, and not a startup in the usual sense, but node-webkit / NW.js-type runtime work. This matters because it explains why he was able to build Electron later at GitHub: he did not simply evolve from being a front-end page author into a framework creator; he entered early into the specialized technical space of integrating a browser engine, Node.js, and native system APIs into a desktop runtime. Kevin Sawicki’s talk even reflects that GitHub initially tried porting Atom to node-webkit, struggled with that route, and then shifted plans. In other words, Cheng Zhao’s entry into Atom/Electron’s core was not accidental; it happened because he was already standing at a narrow but strategically crucial technical intersection.

5、If the subject is the “single-founder myth,” the most accurate summary is this: Cheng Zhao was the key original author and early technical backbone; Kevin Sawicki became an important public face for project history, naming, and outward narrative; and Zeke Sikelianos later helped launch Electron, internationalize the website, add TypeScript support, and cultivate the community. Zeke’s own CV states that during his five years at GitHub, he “helped launch Electron.” So if the question is “who wrote the early technical skeleton,” the answer points primarily to Cheng. If the question is “who turned Electron into a large durable project,” the answer is necessarily plural.

1、Electron’s project lineage is clear. At first it was Atom’s underlying shell; in 2015 it was renamed and began absorbing outside users as an independent framework; by the 2016 1.0 release, the project had already expanded from “a runtime” into a tooling ecosystem, rolling out documentation, API demos, the Devtron debugging extension, the Spectron integration testing framework, and a dedicated community page. That shows Electron was never only trying to sell a runtime; it was building a full developer production system.

2、At the level of brands, assets, organizations, and platforms, Electron’s core assets are not traditional corporate equity. They are open-source infrastructure and technical brand capital: the electron/electron repository, the official site and docs, the release cadence and upgrade machinery, Electron Fiddle, Electron Forge, the community pages, the governance repository, the Open Collective funding channel, and the trademark/governance framework under the OpenJS Foundation. Some of these are engineering assets in a hard operational sense; others are influence assets in an ecosystem sense.

3、Electron Fiddle and Electron Forge explain a large part of why the project kept expanding. Fiddle lowers the barrier for learning, experimentation, and quick prototyping. Forge handles packaging, signing, installers, and publishing in a more unified build pipeline. The official docs explicitly describe Forge as a packaging and publishing tool that unifies Electron’s build ecosystem into a single interface. In other words, Electron succeeded not only because it could build desktop apps, but because it surrounded the entire path from learning to shipping.

4、Cheng Zhao’s own project portfolio is also highly revealing because it forms his personal technical brand. His current website lists Electron, Yue, Chie, node-mlx, and Compilets under “Created,” MLX under “Working on,” and archives including Wey, Crossclip, node-pathwatcher, and asar. The continuity is striking: he repeatedly works on runtimes, GUI tooling, packaging, native bindings, and bridges between JavaScript and system capabilities. Later, the same line extends naturally into local LLM tooling and machine learning infrastructure. Electron is therefore not an isolated career event for him; it is part of a long-running engineering logic.

5、In terms of capital and organizational structure, Electron does not have the classic “startup–VC–funding round” story. It originated at GitHub, but later intentionally moved away from single-company control toward foundation-based neutral governance. In 2019, Electron formalized governance with multiple working groups; the same year it entered OpenJS Foundation incubation; in 2020 it became an OpenJS Foundation Impact Project. The official explanation is explicit: one goal was to lift Electron out of being owned by a single corporate entity and into a neutral foundation maintained by multiple organizations and individuals.

6、But neutral governance does not mean an absence of corporate power. Official documentation today is equally explicit that Electron is maintained by multiple companies, naming Microsoft, Slack/Salesforce, Notion, and others. Cheng Zhao’s 2019 talk also mentioned Microsoft security experts reviewing code and developers from Slack, VS Code, and Microsoft Teams helping with Chromium upgrades. So Electron’s real resource network is not VC capital. It is corporate engineering labor, foundation governance, community contributions, and a modest public funding channel.

7、On the public funding side, Electron’s Open Collective page states that general funds go to contractors, bounties, security audits, performance work, and general upgrades. In other words, Electron does have a public fundraising mechanism, but the function of that money is infrastructural maintenance, not shareholder-style growth storytelling. For an MIT-licensed framework, that is the shape of a mature open-source finance model.

1、Electron’s business model is fundamentally one of creating value by removing cross-platform engineering cost rather than charging license fees. Electron is free, MIT-licensed, and available for commercial use. The entities that usually monetize are not Electron itself, but the app companies, tooling vendors, consultants, and labor markets built on top of it. For GitHub, its direct value was enabling products like Atom to be built faster. For companies adopting it later, its value was allowing a largely web-oriented team to ship multi-platform desktop clients. For maintainers and key contributors, the value converted into career capital, technical authority, and ecosystem influence.

2、One major turning point was the move from “internal tool” to “public framework.” The 2015 rename was the first step; the 2016 1.0 release was the second. Version 1.0 mattered not because the number got bigger, but because the project explicitly framed itself as a mature framework with API stability, better docs, new tools, and support for apps that could feel native on Windows, Mac, and Linux. That was also the period when Electron systematically built debugging, testing, educational, and community infrastructure.

3、A second major turning point was making “keeping up with Chromium” into project discipline. Electron moved to a 12-week major cadence in 2019 and then to an 8-week cadence in 2021. The official explanation directly cites two outside pressures: Chromium’s faster release cycle and the Microsoft Store’s freshness requirements for Chromium-based apps. This decision was critical because it transformed Electron from “a desktop framework that sometimes lags the browser engine” into “a desktop framework whose competitive advantage includes a disciplined upgrade velocity.”

4、A third turning point was the formalization of governance and neutralization of ownership. The 2019 working-group model and the 2020 OpenJS Foundation impact-project transition reduced the structural risk of Electron depending too heavily on one company or a tiny number of individuals. In its ten-year retrospective, the project itself emphasized that governance was one of the reasons Electron could keep operating across many companies, time zones, and maintainers. For a framework that must keep up with Chromium and Node security updates, that is not cosmetic—it is existential.

5、A fourth turning point was the tightening of default security posture. The official security documentation repeatedly stresses that Electron is not a browser and that executing untrusted content presents severe risk. To reduce misuse at scale, Electron made nodeIntegration disabled by default starting in 5.0, contextIsolation enabled by default starting in 12.0, and renderer sandboxing enabled by default starting in 20.0. This shift shows that the project recognized a hard truth: if high-privilege capabilities remained open by default, Electron would eventually be damaged by its own flexibility.

6、A fifth turning point was timely adaptation to Apple platform realities. Electron 11 added Apple Silicon support, and the project has long maintained a Mac App Store submission guide. That matters for Electron’s place in the Mac ecosystem: although it is often criticized as “not native enough,” it is not surviving by avoiding Apple’s platform rules. It has continuously entered Apple’s practical world of signing, store submission, distribution, and chip transitions.

7、Its most impressive result is that Electron turned “web technologies for desktop software” from a suspicious idea into a normal industrial path. Official 2026 docs already describe it as an enterprise-grade framework for flagship products and list ChatGPT, Claude, VS Code, Slack, Discord, Signal, Notion, Docker, Canva, and Loom as real applications. One may still argue that Electron is not always the technically optimal choice, but it is difficult to deny that it has become one of the most powerful and reality-shaping choices.

8、For Cheng Zhao personally, the most significant success is not only that he created Electron, but that he established a durable engineering identity: someone who solves hard problems at the boundary of JavaScript, native systems, GUI, packaging, and runtime integration. That identity later extended into Apple machine learning work and projects such as node-mlx and MLX. In career terms, he did not merely gain one famous project; he built a long transferable technical line.

1、The most persistent negative criticism of Electron is that it is heavy, memory-hungry, slow to start, and less native-feeling than platform-native apps. Interestingly, the project does not deny this. The performance guide opens by acknowledging that developers frequently ask how to reduce memory, CPU, and disk use; the Why Electron page also openly admits that bundling Chromium increases disk footprint and says many Electron apps are larger than 100MB. The official position is not “these costs do not exist,” but “these costs buy compatibility control, security patchability, and cross-platform consistency.”

2、A second major criticism is the security burden. The official security guide repeatedly emphasizes that Electron is not a browser; if remote untrusted content is mixed with Node capabilities, an XSS can easily become an RCE. That is why Electron maintains a highly detailed security checklist: only load secure content, never enable Node integration for remote content, enable context isolation, use the sandbox, validate IPC senders, and restrict navigation and new windows. The right summary is therefore not “Electron is insecure,” but “Electron is extremely powerful, so misuse can be extremely dangerous.”

3、A third area of controversy is that Electron has in fact faced multiple high-severity vulnerabilities. The official security blog records the 2018 Windows protocol-handler remote code execution issue, the 2018 webview-related nodeIntegration bypass, the 2018 webPreferences child-window RCE issue, and later high-severity Chromium-derived vulnerabilities. In other words, Electron, as a large runtime blending Chromium, Node, and native capabilities, naturally lives under heavy security-maintenance pressure. Its strength is not that nothing goes wrong, but that it has institutionalized patching and response.

4、A fourth controversy is more internal to the developer world, but crucial for understanding the founder narrative: the disputed history of NW.js versus Electron. In his 2016 public statement, Roger Wang objected to Cheng portraying node-webkit as essentially a “solo developer” project and stressed that Cheng was an intern he recruited in 2012, and that important Node+Chromium integration existed before Cheng joined. At the same time, Cheng’s current personal site still describes his NW.js work as “Rewrote the project from scratch.” Therefore, on questions such as “who inherited what from whom?” and “was Cheng the formative creator of NW.js or a major later contributor?”, the best possible wording is simply: public accounts differ.

5、Another point often misunderstood is that Atom’s decline does not mean Electron failed. GitHub announced Atom’s sunset in 2022, and the Atom repositories were later archived. Meanwhile, Electron itself continued growing and now covers a far broader spread of applications than it did in the Atom era. In other words, Electron not only outlived its original “parent application,” it escaped Atom’s fate entirely. That is strong evidence that Electron became a platform rather than remaining a product attachment.

6、As for current status, Electron is still actively evolving. The official repository shows v42.3.0 as the latest stable release; documentation states that major releases follow an 8-week cadence tied to Chromium, and that only the latest three stable release lines are officially supported. Current platform support includes macOS Monterey and newer, Windows 10 and newer, and Linux; macOS binaries are provided for both Intel and Apple Silicon. This means Electron today is not a historical artifact. It is a mature infrastructure project under active, high-frequency maintenance.

7、Cheng Zhao’s present role is also clear. He currently works at Apple on machine learning, and his website shows MLX as his present focus while he continues maintaining a collection of projects around JavaScript, local models, native bindings, and runtime tooling. In practical terms, he is no longer simply “the framework author who still lives off Electron,” but rather “the systems engineer who created Electron and has since moved into Apple machine-learning infrastructure.” Electron remains his signature work, but it is not his endpoint.

8、If the entire story is compressed into a short timeline, it looks like this: around 2012, Cheng Zhao enters the node-webkit / NW.js context; in 2013 Atom Shell begins; in 2014 Atom enters public beta; in 2015 Atom Shell is renamed Electron; in 2016 Electron 1.0 arrives and the docs/tooling ecosystem takes shape; in 2019 governance is formalized; in 2019–2020 the project moves into the OpenJS Foundation; in 2021 the major-release cadence becomes 8 weeks; by the mid-2020s Electron has become one of the mainstream desktop shells of the AI-and-collaboration software era, while Cheng Zhao himself has extended his long-term strength—bridging JavaScript and low-level systems—into Apple machine learning work.