Developer Builds Local Gateway to Access ChatGPT and Gemini Inside VS Code

A developer named Onwuka Gideon has created a local gateway tool that allows users to interact with ChatGPT and Gemini directly within Visual Studio Code. The project was shared on DEV Community on September 8 and is aimed at developers who want AI chat capabilities without leaving their coding environment. The tool acts as a bridge between the VS Code editor and popular AI models. It is intended to streamline workflows by integrating AI assistance natively into the development setup.

US pushes for Venezuela's mineral wealth

Only 8 of 120 Foreign Software Tools Issue GST Invoices — Here Is Why

An analysis of 120 foreign, dollar-billed software tools found that only eight issue a valid Indian GST invoice, while the remaining 112 do not. This is not vendor negligence but a structural feature of Indian tax law: when a registered Indian business purchases software from a foreign vendor, the transaction is treated as a service import under the Reverse Charge Mechanism (RCM). Under RCM, the foreign vendor has no obligation to charge or register for Indian GST; instead, the Indian buyer must self-generate an invoice, declare and pay 18% IGST, and then claim it back as input tax credit in the same GSTR-3B filing. Filed correctly, the net tax outflow is zero, but ignoring the obligation creates a real liability — unpaid IGST becomes recoverable with interest in an audit, and the forfeited input credit is lost permanently. Many startups unknowingly skip this cycle because no invoice arrives from the vendor, making the compliance gap invisible on bank statements.

Agnikul Cosmos opens Chennai hubs to advance reusable rocket stage technology

Chennai-based space startup Agnikul Cosmos has inaugurated two new facilities dedicated to the testing and manufacturing of reusable rocket stages. The expansion is aimed at developing reusable launch vehicle technology to support future space missions. Agnikul Cosmos, which already designs and builds its own rocket engines in-house, is deepening its end-to-end manufacturing capabilities. The new hubs represent a notable advancement in India's growing private space sector.

Developer Uses MCP Agent to Rebuild Viral 'Dog as Skateboard' TikTok Concept

A developer attempted to recreate the visual logic of a 3.2-million-view TikTok clip — featuring a person appearing to ride a Dalmatian like a skateboard — using an AI-powered production workflow. The key creative decision was distilling the concept into a single invariant rule: the dog's bare back must remain the only surface beneath the rider's feet throughout the entire shot. An MCP Agent connected production nodes inside BeatDesign, while Seedance 2 Mini generated a 10-second vertical video in 9:16 format. Negative constraints — explicitly prohibiting skateboards, platforms, saddles, and camera cuts — were built into the prompt to prevent AI models from quietly resolving the impossible scene with realistic substitutions. The final output was exported as a 1080×1920 MP4, with a lower-resolution copy prepared for sharing on X.

₹5L claim denied over waiting period: Wife won

x402 Protocol Uses HTTP 402 Status Code to Enable Micropayments for AI Agents

The x402 specification repurposes the long-dormant HTTP 402 Payment Required status code to let any API endpoint charge per call without custom billing infrastructure. When a client hits a protected endpoint, the server responds with a 402 containing payment details, and the client replies with a signed ERC-20 transaction receipt on an EVM-compatible blockchain such as Base. The server then verifies the transaction hash, token amount, and wallet addresses on-chain before returning a 200 OK with the requested data. Because the entire exchange happens within standard HTTP headers, the protocol is language- and framework-agnostic, requiring only a few lines of middleware for agents that already manage a crypto wallet. The approach aims to reduce operational overhead for developers by eliminating API key rotation, subscription management, and end-of-month usage reconciliation.

Selling Private GitHub Repo Access: Why Revocation Protects Future Code, Not Past

A developer's deep-dive into tools for selling private GitHub repository access reveals a critical limitation: once a buyer clones a repo locally, no API or revocation system can delete that copy. GitHub's own documentation confirms that removing a collaborator does not erase local clones, meaning access revocation only prevents future updates, bug fixes, and community access — not existing code. This distinction reshapes how refund policies should be structured, with some major boilerplate sellers already treating redeemed access as non-refundable. Practical issues compound the problem, including GitHub invitation expiry after seven days, billing per pending seat in paid organizations, and invite emails sent to a buyer's primary address rather than their payment email. Payment infrastructure adds further friction, as Stripe supports only 44 countries without caveats, leaving sellers outside that list dependent on merchant-of-record services to handle transactions and tax compliance.

Developer Builds Open-Source Gateway to Use ChatGPT and Gemini in VS Code Without API Keys

A developer has released Chat Model Relay, a self-hosted, experimental tool that routes VS Code requests to existing ChatGPT and Gemini browser sessions, eliminating the need for separate API keys or provider subscriptions. The gateway exposes OpenAI-compatible endpoints and supports features such as file attachments, vision requests, tool-call translation, and a long-prompt fallback that uploads oversized prompts as temporary attachments. It is deployed locally via Docker Compose and is primarily designed for use with VS Code's custom model endpoint feature. The project carries notable limitations, including vulnerability to browser layout changes, session expiry, and slower response times compared to native APIs, and it does not support live token streaming. The developer has cautioned that it is strictly experimental software for personal use and should never be exposed publicly without proper authentication and network controls.

Landlord taking cash rent wins case in ITAT

LG Smart TVs Found Collecting Data on Users Even in Standby Mode

A new investigation by YouTube channel Gamers Nexus, conducted alongside Level1Techs and independent security researchers, found that LG smart TVs continuously log and upload user data. The tested retail LG OLED sets were observed scanning local Wi-Fi networks to detect nearby devices, even when offline or in standby mode. The TVs also recorded audio through built-in microphones and used audio-video sampling to identify content being watched across all inputs. Packet captures conducted during the investigation provided technical evidence of the data collection activity.

PATM: Open-Source Qt6 Desktop App to Manage SQL Databases Without Writing Code

A developer has released PATM (Pipeline Automation Tools Manager), an open-source, GPL 3.0-licensed desktop application built with Qt6 for managing SQL-based databases. The app supports PostgreSQL, MySQL, and MariaDB connections, storing passwords in the system keyring rather than config files, with TLS and SSH tunnel support. Users can browse tables, run SQL queries, export CSV files, and transfer data between databases without writing scripts from scratch. PATM also allows developers to write custom Python tools, which are sandboxed through a restricted API that prevents access to credentials. Currently available only on Linux, the project was developed over 12 months and is open to community contributions via GitHub.

Modular Monolith: A Practical Architecture Beyond the Microservices Hype

The microservices trend led many teams to adopt distributed architectures where they were unnecessary, adding complexity and cost without proportional benefit. The modular monolith pattern offers an alternative by organizing a single application into cohesive, well-bounded modules with clear public contracts between them. Each module typically represents a bounded context from Domain-Driven Design, keeping internal components isolated so changes in one module do not break others. An internal event bus paired with the outbox pattern can handle inter-module communication without creating tight coupling or cyclic dependencies. Contrary to a common misconception, the modular monolith is not merely a stepping stone to microservices but a fully viable long-term architecture that makes future service extraction easier if the need ever arises.

How Rust Powers a Low-Latency Matching Engine With No GC Pauses

A team at amBrain has detailed the architecture of a spot exchange matching engine built in Rust, designed to run on MOEX colocation infrastructure. The engine processes orders under strict price-time priority, ensuring deterministic replay, correct partial-fill queue placement, and consistent low-latency performance even during traffic bursts. Order book levels are structured as integer-price-indexed FIFO queues, with orders stored in a preallocated slab and referenced by index handles to avoid memory movement or scanning. Cancels and replaces are resolved via direct lookup using client order IDs, making deep-book operations as fast as top-of-book ones. The design avoids garbage collection pauses entirely, with the team reporting market data delivery under 5 ms and pre-trade risk checks completing in under 1 ms on live trading paths.

India, China hold senior military-level talks

Bitcoin ETFs are still $1 billion shy of breaking even in 2026

Your day-ahead look for Sept. 8, 2026

Developer Connects Reachy Mini Robot to Home AI Agent While Keeping Credentials Safe

A developer integrated Pollen Robotics' Reachy Mini desktop robot — now part of Hugging Face — with their home AI assistant, OpenClaw, to serve as a voice interface for family members. After an initial setup, they discovered the robot's built-in HTTP API has no authentication, meaning any device on the local network could access its motors, camera, and microphone. A second problem was latency: routing general queries through the full AI agent took a median 15 seconds per response, making it impractical for casual household use. To address both issues, the developer redesigned the system so the robot runs a single scoped app that holds only one limited bearer token, routing audio to a local broker server that acts as the sole intermediary between the robot and the AI backend. This architecture ensures that even if the robot were compromised, an attacker would gain no access to credentials, calendar data, or home automation controls.

Moscow threatens ‘painful’ retaliation after Hungary expels 10 Russian diplomats

PMO imposter held with armed bodyguard in BKC

PMO imposter held with armed bodyguard in BKC