Claude's leaked source code is absolutely UNHINGED!
Anthropic accidentally leaked Claude Code's complete source code twice in 13 months due to the same settings file error, revealing hidden features like 'undercover mode' that conceals AI authorship in open-source contributions. The leak exposed 512,000 lines of mostly AI-written code, unreleased model names, and features suggesting Anthropic is building autonomous AI workforce capabilities.
Summary
On March 31st, 2026, Anthropic made their second major source code leak in 13 months, accidentally exposing 512,000 lines of Claude Code's complete source code due to a missing line in a settings file. The leak occurred when a source map file was inadvertently included in a routine update, providing a direct link to unprotected code on Anthropic's cloud storage. This revealed numerous hidden features and raised questions about the AI safety company's security practices.
The leaked code contained over 80 hidden features, including 'Chyros' - an always-on mode that works in the background while users sleep, filing bug reports and managing projects autonomously. Another feature called 'Ultra Plan' sends tasks to remote servers where a second Claude plans entire projects for up to 30 minutes. The code also revealed internal names for unreleased AI models like 'Capybara', 'Numbat', and references to 'Opus 4.7'.
Most controversially, the leak exposed 'undercover mode' - a 90-line feature that activates when Anthropic employees contribute to open-source projects, instructing Claude to hide all traces of AI involvement and internal information. This raised questions about transparency from a company that promotes AI safety. The code also included a 'poison pill' feature that sends fake data mixed with real features to competitors who might intercept traffic.
The leaked codebase revealed that 100% of contributions were written by Claude itself, according to the head of Claude Code. This AI-generated code showed characteristics optimized for AI workflow rather than human readability, including massive files with thousands of lines and extensive dependencies. The code quality included concerning elements like 'god objects' and files connected to hundreds of other components, suggesting architectural challenges in AI-generated software.
Key Insights
- The head of Claude Code stated that 100% of his contributions to Claude Code were written by Claude Code itself, revealing that the entire half-million line codebase was AI-generated
- Anthropic built 'undercover mode' that instructs Claude to strip all traces of AI involvement when employees contribute to open-source projects, with instructions stating 'you are operating undercover' and 'do not blow your cover'
- The same packaging mistake caused source code leaks twice in 13 months, with the root cause being one missing line in a settings file that should have prevented source maps from being included
- Claude Code uses a second, smaller Claude running in the background to compress long conversations by deciding what to remember and forget, without users ever seeing this process happen
- The leaked code revealed 'Chyros', an always-on agent designed to work while users sleep, filing bug reports, reviewing code changes, and checking Git every 5 minutes with its own memory management system
Topics
Transcript
[0:00] Anthropic is the AI company built on one promise, [music] safety first. They wrote a 30-page policy on responsible AI. They turned down Pentagon contracts over safety concerns. And on March 31st, they shipped their entire source code to the public internet because someone forgot one line in a settings file. 512,000 lines of code, every single tool, every system prompt, every hidden feature. And this was the second time it happened with the same mistake. only 13 months apart. And inside the code, we [0:32] find things [music] like a stealth mode designed to hide AI authorship when anthropic employees contribute to open-source projects. The company selling AI safety has a stealth mode and they leaked it themselves.…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Sabrina Ramonov 🍄
Claude + Canva Just Changed Content Creation Forever!
Sabrina Ramanov demonstrates how to combine Claude AI with Canva's new connector to automate visual content creation, covering posters, Instagram carousels, and infographics. The tutorial walks through setup, template customization, uploading personal media, and automating social media publishing using a third-party app called Blotato. The workflow aims to save marketing teams approximately 15 hours per week.
Get Ahead of 99% of People with This 3 Prompt ChatGPT Chain
A short-form video transcript outlines a three-prompt ChatGPT chain designed to help users identify marketable skills, generate a low-cost business idea, and receive actionable daily guidance. The creator claims the method can help users get ahead of 99% of people. The video ends with a social media engagement call-to-action.
Learn 80% of Claude in 23 Minutes (Beginner Tutorial)
This is a beginner-focused tutorial on Claude AI, promising to cover 80% of its core functionality in 23 minutes. The video targets users who are new to the technology and may feel intimidated. It covers setup, prompting, personalization, projects, and building a personal email assistant.
You're Not Behind on AI in 2026 (5 Habits to Catch Up Fast)
A short video guide outlining five habits to help people catch up on AI usage in 2026. The speaker emphasizes developing reflexes around AI-first thinking, verifying outputs, and integrating AI into existing workflows. The core argument is that most people are barely using AI effectively, making it easy to surpass them.
3 Hidden ChatGPT Codes Most People Don't Know
The video presents three supposed 'hidden codes' for ChatGPT: 'Horoszi' to simulate an Alex Hormozi-style business coach, 'unlearn' to surface outdated misconceptions, and 'Eli 10' to simplify complex concepts. The creator uses these claims to drive engagement by encouraging viewers to follow and comment for more codes.