AhmadHassan-BTed/Noria

The application can now be containerized using Docker. A production Dockerfile is provided that builds the Node.js application, installs dependencies, and runs as a non-root user with a health check. A development Dockerfile is also added, which includes hot-reloading via nodemon. A docker-compose configuration is introduced to manage both production and development services, mapping ports 3000 and 3001 respectively, and mounting necessary volumes for authentication and logs.

docker · high confidence

Added a new example file, examples/custom-provider.js, which demonstrates how to create a custom provider for job opportunities. This example includes a JobAnalyzer that evaluates job postings based on skills, salary, and experience, and a JobNotifier that formats the results. The provider exports JobsProvider, JobAnalyzer, JobNotifier, and the response schema, illustrating how to extend the system with custom logic.

examples · high confidence

The build system now generates the necessary configuration files (Analysis, EXE, PKG, PYZ, and warnings) to package the \noria\_launcher\ application into a standalone executable (\Noria.exe\). This change establishes the foundation for distributing the application as a self-contained Windows binary, incorporating standard Python libraries and system dependencies required for the launcher to function without a pre-installed Python environment.

build · high confidence

The src/utils directory now includes five new modules: cache.js provides an LRU cache with TTL and size limits; metrics.js tracks scraper, analyzer, WhatsApp, and LLM (Gemini/Groq) events with a formatted report; queue.js implements a dead-letter queue with exponential backoff and retry scheduling; retry.js offers a robust retry wrapper that parses rate-limit headers and handles network errors; and validators.js enforces URL, scraper payload, and analyzer response schemas. These utilities support the new pipeline orchestration and resilience features.

src/utils · high confidence

Added new domains for processing job and scholarship opportunities using Large Language Models. The \src/domains/jobs\ and \src/domains/scholarships\ directories now contain the core logic for evaluating opportunities, including scoring criteria, response schemas, prompt builders, and message templates. To support these pipelines, the \src/infrastructure/llm\ directory introduces a stateless multi-provider dispatcher (\fallback.js\) that routes requests to a configurable chain of LLM providers (e.g., Google Gemini and Groq). The implementation includes specific adapters for each provider (\gemini.js\, \groq.js\) and a request queue (\requestQueue.js\) to manage API rate limits.

src/infrastructure · high confidence

The application now features a new event-driven pipeline architecture, bootstrapped via a new \src/index.js\ entry point that initializes a \PipelineOrchestrator\ and loads pipeline configurations from YAML files. Additionally, \src/launcher.js\ has been added to provide a dynamic launch interface for pipeline instances, featuring global console output redirection to session-specific log files and robust error handling for uncaught exceptions and unhandled rejections.

src · high confidence

The application's frontend has been refactored from a monolithic structure into a modular \ui/views\ package, separating concerns into distinct view files. This change introduces a dedicated profile management interface that allows users to configure multi-provider LLM chains (supporting Gemini and Groq) and Jina Reader keys. It also adds a scans management view for launching and monitoring opportunity extraction pipelines, and a settings view for managing Windows startup integration and performing a full application data reset. These components collectively enable users to manage applicant profiles, configure AI backends, and control background scanning operations.

ui/views · high confidence

The \ui/\ directory is introduced, separating frontend logic into \state.py\ and \styles.py\. \state.py\ manages application state including predefined scan templates, profile persistence, and session logs. \styles.py\ applies a comprehensive dark theme using CSS variables (e.g., \--bg-page: \#212121\) and customizes Streamlit components, typography, and layout spacing to match a WhatsApp-inspired aesthetic.

ui · high confidence

The application now uses a new core architecture that manages data processing pipelines through a central orchestrator. This system reads YAML configurations to wire together distinct stages—such as listeners, scrapers, LLM analyzers, and senders—using a function-based registry. Users can now define and run automated workflows (e.g., for scholarships or jobs) where each step is a decoupled module, allowing for more flexible and extensible data processing.

src/core · high confidence

The scripts directory now includes a suite of new automation tools: build.js and build\_exe.py handle the build process, including compiling a standalone C\# launcher (Noria.exe) with a custom icon; launcher.cs provides a cross-platform startup experience that checks for Python and Node.js dependencies and launches the application in an app window; publish\_release.py automates GitHub release creation with the compiled executable as an asset; and supporting scripts (convert-ico.py, embed-logo.py, render-logo.js, test-models.js) handle asset conversion, embedding, and API testing.

scripts · high confidence

Noria v1.0.0 launches as a decoupled, hexagonal pipeline orchestrator for opportunity extraction and generative evaluation. The update introduces a Streamlit-based control center (app.py) that manages profiles, scans, and settings. It supports multi-LLM fallback chains (Gemini and Groq) for analysis, alongside WhatsApp-based notifications and scanning. The release includes a Makefile for development workflows, a Windows launcher (run\_noria.bat) that opens the app in a dedicated Chrome/Edge window, and a Dockerfile for containerized deployment. Configuration is standardized via a new .env.example template, and the project adopts the MIT license.

(repo-wide) · high confidence

The application now features a new event-driven architecture with a dedicated queue broker. Users will see a formatted system error log that displays the error source, URL, and message in a structured box, along with a boot banner confirming the pipeline is live. This replaces previous, less structured console output for system events.

src/config/constants, src/queue · high confidence

The monolithic frontend code has been refactored into a modular \ui/components\ package, separating concerns into distinct files for countries, device cards, and navigation. This change introduces a new sidebar navigation component that provides a unified dashboard for managing applicant profiles, opportunity scans, and system settings. The device card component has been updated to handle connection linking, QR code scanning, and status tracking for WhatsApp-linked devices.

ui/components · high confidence

Add unit tests for LLM, messaging, and domain modules; Added unit tests for core event, pipeline, and registry modules; Added unit tests for utility modules.

The project's dependency tree has been updated. The \package-lock.json\ file was regenerated, and the \requirements.txt\ file was added to manage Python dependencies. Additionally, several \node\_modules\ directories were removed from the repository, indicating a cleanup of vendored or cached dependencies.

(dependencies) · medium confidence

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

• CAI 64 → 61 (-3.4)
• Rubric changed (rubric-2026.08.18 → rubric-2026.09.15) — scores are not directly comparable.

• Code Health 58 → 48 (-10.4)
• Architecture 99 → 85 (-13.2)
• Maturity 75 → 76 (+1.2)
• Readiness 61 → 67 (+6.6)
• Security 68 → 76 (+8.0)

• Change coupling: app.py ↔ whatsapp-listener.js (app.py)
• Coverage not included — suite not readable by the collector
• Dependency hygiene not measured — dependency manifest found but not parsed for hygiene
• Duplicated block (17 lines × 2) (ui/views/profiles.py)
• Duplicated block (23 lines × 2) (ui/views/profiles.py)
• Duplicated block (31 lines × 2) (ui/views/profiles.py)
• Duplicated block (6 lines × 2) (ui/components/device_card.py)
• Duplicated block (9 lines × 2) (ui/components/device_card.py)
• Duplicated block (9 lines × 2) (ui/views/profiles.py)
• High CVE: [GHSA redacted] (package-lock.json)
• High CVE: [GHSA redacted] (package-lock.json)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• High: security finding (details withheld)
• …and 7 more

• ChannelFetcher._installBridge (cyclomatic 42) (src/infrastructure/messaging/channel-fetcher.js)
• ChannelFetcher._strategyNewsletterStore (cognitive 23) (src/infrastructure/messaging/channel-fetcher.js)
• ChannelFetcher._strategyNewsletterStore (cyclomatic 19) (src/infrastructure/messaging/channel-fetcher.js)
• ChannelFetcher.fetchWithRetry (cognitive 17) (src/infrastructure/messaging/channel-fetcher.js)
• ClassTooLong: WhatsAppListener (src/infrastructure/messaging/whatsapp-listener.js)
• Coverage not measured — JavaScript/TypeScript suite
• Dependency advisory scan runs only on code events
• Dependency hygiene PARTLY measured — Python dependencies read, no exact pin to grade for currency
• Documentation: no installation or build instructions (README.md)
• Duplicated block (10 lines × 2) (ui/components/device_card.py)
• Duplicated block (11–16 lines × 3) (ui/views/profiles.py)
• Duplicated block (22–26 lines × 3) (ui/views/profiles.py)
• Duplicated block (35–36 lines × 2) (ui/views/profiles.py)
• Duplicated block (5 lines × 2) (ui/components/device_card.py)
• Duplicated block (8 lines × 2) (ui/views/profiles.py)
• FileTooLong: messaging/whatsapp-listener.js (src/infrastructure/messaging/whatsapp-listener.js)
• High CVE: [GHSA redacted] (package-lock.json)
• High CVE: [GHSA redacted] (package-lock.json)
• High CVE: [GHSA redacted] (package-lock.json)
• High CVE: [GHSA redacted] (package-lock.json)
• …and 69 more

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

AhmadHassan-BTed/Noria was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.