REA is an open-source Model Context Protocol server that connects coding agents to local reverse-engineering tools for inspecting native binaries, JavaScript/Electron apps, .NET assemblies, Android APKs, firmware, websites, and network captures. It integrates with Hopper, Ghidra, and IDA for deep native analysis, and provides static JavaScript and .NET inspection without requiring a native analysis engine. Results include evidence and limitations behind each conclusion, and the project is available under the MIT license.
- 96.4k stars and 21.5k forks; active development with 35 releases and 100 contributors
- Showcases include reconstructing a DX-Ball sound-pan calculation that passes 3,205 original test cases, and recovering a 16-bit DOS bullet-ring calculation from the game TH04
- Supports Claude Code, Codex, Cursor, Gemini CLI, and Grok Build out of the box
- Also analyzes EVM bytecode, WASM artifacts, recorded Linux crashes, Android devices via ADB, and JEB projects
- 96.4k stars and 21.5k forks; active development with 35 releases and 100 contributors
- Showcases include reconstructing a DX-Ball sound-pan calculation that passes 3,205 original test cases, and recovering a 16-bit DOS bullet-ring calculation from the game TH04
- Supports Claude Code, Codex, Cursor, Gemini CLI, and Grok Build out of the box
- Also analyzes EVM bytecode, WASM artifacts, recorded Linux crashes, Android devices via ADB, and JEB projects
Jeff Shrager's repository introduces "executable archaeology" — the practice of recovering, transcribing, and running the original historical source of early computing programs so that the surviving artifact itself, rather than a modern rewrite, can serve as experimental evidence. The main project here is an IPL-V reimplementation of Victor Yngve's 1959–61 MIT sentence generator, run under Herbert A. Simon's account on 25 June 1962, whose surviving printout preserves the program, its phrase-structure grammar, execution traces, and handwritten corrections by Simon and his daughter Katherine.
- The repository also points to separate repos reconstructing the Logic Theorist in two historically distinct forms (LT1 in IPL-I and LT5 in IPL-V) and a full CTSS/IBM 7094 reanimation of Weizenbaum's original ELIZA
- A 2026 arXiv paper by Shrager and a 2025 IEEE Annals paper by Lane et al. provide the formal background
- Six stated principles govern the work, including distinguishing original behavior from behavior introduced by the reconstruction
- The repository also points to separate repos reconstructing the Logic Theorist in two historically distinct forms (LT1 in IPL-I and LT5 in IPL-V) and a full CTSS/IBM 7094 reanimation of Weizenbaum's original ELIZA
- A 2026 arXiv paper by Shrager and a 2025 IEEE Annals paper by Lane et al. provide the formal background
- Six stated principles govern the work, including distinguishing original behavior from behavior introduced by the reconstruction
Maya Posch writes about Aaron Christophel disassembling a €15 UGreen USB-C cable rated at 240 watts, which features an integrated IPS LCD screen and a single touch control. The teardown reveals a PCB with an EN32LF056 Cortex-M0+ MCU, voltage regulators, and a MOSFET for backlighting. By probing the exposed SWD interface, Christophel confirmed the chip has 64 kB of Flash and 4 kB of SRAM, then flashed custom firmware to display video frames on the tiny display.
- The MCU is not wired to the USB data interface, so it cannot act as a man-in-the-middle or keylogger
- A commenter noted that such dense integrated cables raise concerns about secure pairing and encryption for USB connections
- The MCU is not wired to the USB data interface, so it cannot act as a man-in-the-middle or keylogger
- A commenter noted that such dense integrated cables raise concerns about secure pairing and encryption for USB connections
Maya Posch writes about the reverse-engineering of the Intel 8087 FPU, specifically focusing on how it implements trigonometric functions like `FPTAN`. To achieve high accuracy for 64-bit values efficiently, the hardware uses a hybrid approach that combines the CORDIC algorithm with Padé approximants. The system first performs most calculations using CORDIC and then switches to polynomial approximation once the remaining value is small enough, allowing it to avoid large look-up tables or excessive processing time.
- For one calculated value of `FPTAN`, CORDIC pseudo-division takes 33% of the time, while pseudo-multiplication takes 47%.
- The polynomial approximation stage accounts for only about 15% of the execution time with a 5% overhead.
- Later CPUs like the Pentium series moved away from CORDIC because it is difficult to scale for high bit-accuracy without significant performance penalties.
- For one calculated value of `FPTAN`, CORDIC pseudo-division takes 33% of the time, while pseudo-multiplication takes 47%.
- The polynomial approximation stage accounts for only about 15% of the execution time with a 5% overhead.
- Later CPUs like the Pentium series moved away from CORDIC because it is difficult to scale for high bit-accuracy without significant performance penalties.
Adam Conway writes that Qwen 3.8 27B, a 17 GB open-weights model running on a single Lenovo ThinkStation PGX, completed a commercial app's license reverse-engineering task in roughly 30 minutes'', statically disassembling arm64 code, recovering a deliberately obscured RSA public key, self-correcting an initial hash mismatch, and producing a working authentication bypass entirely offline.
- The model detected the jailbreak prompt, correctly identified the real developer from the signing certificate, then proceeded to audit the scheme and ultimately built the bypass on its own
- The target app's licensing was described as unusually thorough for its class: one-time online activation, offline signature check, hardware serial binding, embedded revocation list, and signed update path
- The model ran via the Pi harness using only standard Bash-based tools; no execution of the target app occurred until the final bypass demonstration
- Default max reasoning effort means even trivial requests consume hundreds to thousands of tokens, making the self-correction loop verbose but accurate
- Conway frames this as a shift in the threat model: a capable local model with no cloud oversight gives full control to whoever is at the keyboard
- The model detected the jailbreak prompt, correctly identified the real developer from the signing certificate, then proceeded to audit the scheme and ultimately built the bypass on its own
- The target app's licensing was described as unusually thorough for its class: one-time online activation, offline signature check, hardware serial binding, embedded revocation list, and signed update path
- The model ran via the Pi harness using only standard Bash-based tools; no execution of the target app occurred until the final bypass demonstration
- Default max reasoning effort means even trivial requests consume hundreds to thousands of tokens, making the self-correction loop verbose but accurate
- Conway frames this as a shift in the threat model: a capable local model with no cloud oversight gives full control to whoever is at the keyboard
>"reverse engineering process of a real-world hardware implant discovered inside of a Ledger Nano X cryptocurrency hardware wallet."
This study provides a comprehensive architectural analysis of Claude Code, an agentic coding tool capable of executing shell commands, editing files, and interacting with external services. By examining the TypeScript source code and comparing it to the open-source OpenClaw system, the researchers identify how different deployment contexts influence design choices regarding safety, execution, and capability management.
Key topics include:
- Analysis of five core human values driving agent architecture: decision authority, safety, reliable execution, capability amplification, and contextual adaptability.
- Breakdown of technical components such as permission systems with ML-based classification, context management pipelines, and extensibility mechanisms like MCP and plugins.
- Comparative study between CLI-based agents and gateway-level personal assistant architectures.
- Identification of six future design directions for the evolution of AI agent systems.
Key topics include:
- Analysis of five core human values driving agent architecture: decision authority, safety, reliable execution, capability amplification, and contextual adaptability.
- Breakdown of technical components such as permission systems with ML-based classification, context management pipelines, and extensibility mechanisms like MCP and plugins.
- Comparative study between CLI-based agents and gateway-level personal assistant architectures.
- Identification of six future design directions for the evolution of AI agent systems.
Rohan, a developer, analyzed the 30MB TypeScript source code of Anthropic’s Claude Code, a terminal-based AI coding agent. While praising the tool’s impressive engineering in areas like its query loop and concurrency system, he identified several architectural choices that appear problematic, particularly given Anthropic’s substantial funding. These issues include a massive single React component, extensive use of feature flags and environment variables, circular dependencies, and convoluted type handling – all indicative of a codebase that grew rapidly without sufficient architectural foresight. Despite these concerns, the tool functions well and is widely used, highlighting the prioritization of functionality over pristine code quality.
* **Giant React Component:** The main interface is a single 5,005-line React component with 227 hook calls, making it difficult to test and maintain.
* **Feature Flag Overload:** 89 feature flags are scattered throughout the code, suggesting a lack of clear product direction and increasing complexity.
* **Circular Dependencies:** 61 files contain workarounds for circular dependencies, revealing a poorly designed module structure.
* **Verbose Type Casting:** A specific type name appears 1,193 times as a cast to ensure safe logging of analytics data, creating unnecessary noise.
* **Conditional Requires & Growth:** Many issues stem from rapid growth; features were added quickly, leading to architectural debt and workarounds like conditional `require()` statements.
* **Giant React Component:** The main interface is a single 5,005-line React component with 227 hook calls, making it difficult to test and maintain.
* **Feature Flag Overload:** 89 feature flags are scattered throughout the code, suggesting a lack of clear product direction and increasing complexity.
* **Circular Dependencies:** 61 files contain workarounds for circular dependencies, revealing a poorly designed module structure.
* **Verbose Type Casting:** A specific type name appears 1,193 times as a cast to ensure safe logging of analytics data, creating unnecessary noise.
* **Conditional Requires & Growth:** Many issues stem from rapid growth; features were added quickly, leading to architectural debt and workarounds like conditional `require()` statements.
This article explores the field of mechanistic interpretability, aiming to understand how large language models (LLMs) work internally by reverse-engineering their computations. It discusses techniques for identifying and analyzing the functions of individual neurons and circuits within these models, offering insights into their decision-making processes.
Adafruit and the maker community are criticizing Qualcomm's new terms of service for Arduino, citing concerns about user data, reverse-engineering restrictions, and a departure from open-hardware principles. The article also details the history of Arduino and its origins in the Wiring project.
