Open Design turns any coding agent into a design engine, exports real files

·1 views

Open Design, an open-source desktop app from nexu-io, delegates all design work to whichever coding agent a user already has installed, such as Claude Code, Codex, Cursor, or Gemini CLI. Rather than using a proprietary canvas or format, it outputs standard files including HTML prototypes, MP4 motion graphics, PPTX decks, and high-resolution images. Teams can encode their brand into a DESIGN.md file that acts as a brand contract, ensuring consistent fonts, spacing, and components across every generated artifact. The app also ships an MCP server and a BYOK proxy, allowing it to connect to over twenty CLI tools or directly to API endpoints from providers like OpenAI, Anthropic, and Google. The project, which has accumulated around 78,000 GitHub stars, recently released version 0.13.0 and introduced a paid cloud service billed by token usage.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

How Agentic AI Layers Using LangGraph and MCP Can Strengthen CI/CD Security

Traditional CI/CD pipelines excel at automating builds, tests, and deployments but lack the contextual reasoning needed to assess whether a change is truly safe to deploy. Risky operations like unguarded SQL deletions or removed authorization middleware can slip through when no explicit rule exists to catch them. A proposed hybrid architecture addresses this gap by placing a stateful AI reasoning layer — built with tools like LangGraph, MCP, and A2A — above conventional deterministic scanners such as SQLFluff, CodeQL, and Semgrep. This agentic layer collects evidence from specialized tools and evaluates deployment changes from multiple perspectives before producing an auditable decision. The approach is designed to complement, not replace, existing CI/CD security controls, following guidance from NIST and OWASP on integrating security throughout the software development lifecycle.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Fluctara Platform Aims to Adapt Audio Entrainment in Real Time Using EEG Feedback

Fluctara is a newly described audio entrainment platform designed to move beyond static binaural beat tracks by dynamically adjusting audio signals based on a user's physiological state. The system uses a 16-layer coupled-oscillator mathematical framework called the Self-Consistent Phenomenological Network to generate binaural, isochronic, and spatial audio computationally rather than from prerecorded loops. When connected to EEG or heart-rate sensors, the platform operates as a closed-loop system, processing real-time physiological data to modify session parameters for each individual. A core feature called the Entrainment Verification Score attempts to measure how closely a user's recorded brainwave activity correlates with the intended target frequency, going beyond simply logging listening time. The platform is positioned as a tool for researchers, developers, and wellness or human-performance organisations seeking a programmable, verifiable signal-generation system.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Database Migrations: How to Manage Schema Changes Like Code

Database migrations are version-controlled, ordered scripts that evolve a database schema in a traceable, repeatable way, replacing risky manual SQL changes. Tools like EF Core, Flyway, Liquibase, and DbUp track which migrations have run, in what order, and across which environments. Without migrations, teams face environment drift, no audit trail, and no peer-review process for potentially disruptive schema changes. Treating migrations like application code — committed to Git and reviewed via pull requests — ensures dev, staging, and production databases remain consistent. The approach also supports zero-downtime deployments, safe rollback strategies, and seamless integration into CI/CD pipelines.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

AI-Generated GIS Code Can Run Without Errors Yet Deliver Wrong Spatial Results

AI coding agents can generate geospatial code that executes cleanly but produces methodologically incorrect results, a problem distinct from simple API errors. A common example is buffer operations run on datasets using geographic coordinate systems like EPSG:4326, where units are degrees rather than metres, yielding geometries that silently misrepresent real-world distances. Similarly, standard random train-test splits used in spatial machine learning ignore spatial autocorrelation, inflating model accuracy scores that collapse when applied to new regions. Map rendering adds another layer of risk, as visually convincing outputs can mask misregistered rasters, misleading projections, or dropped geometries. Experts argue that defensible GIS workflows require explicit coordinate reference system checks, spatially blocked validation, and domain-aware guardrails that go beyond what general-purpose AI agents currently apply by default.

0 comments Read more at DEV Community