sickn33/agentic-awesome-skills

Agentic Awesome Skills is a curated library of instructional 'skills' designed to guide AI coding agents in performing specific development, security, and operational tasks. It provides a static web application for browsing, comparing, and installing these skills, alongside a comprehensive plugin ecosystem covering domains like API design, accessibility, and security auditing. The system functions as a structured knowledge base that standardizes agent workflows through detailed documentation, code templates, and validation tools.

The 'AAS Agent & MCP Builder' plugin bundle is now available at version 17.5.0, providing a comprehensive set of skills for building and managing agentic applications. This release introduces an 'Agent Evaluation' skill for running versioned behavioral tests and calculating reliability metrics, alongside a 'Langfuse' skill for implementing production-grade LLM observability and tracing. It also includes updated guides for 'LangGraph' stateful agent construction, 'Context Window Management' strategies, and 'LLM Application Patterns' covering RAG and agent architectures. The bundle is distributed via plugin manifests for Claude Code, Codex, and the agent-plugins standard.

plugins/agentic-bundle-aas-agent-mcp-builder · high confidence

Users can now leverage two new agent skills in the skills directory: the '00-andruia-consultant' skill, which acts as a Spanish-language Solutions Architect to diagnose project states and generate implementation roadmaps, and the '007' skill, a critical-risk security audit tool that performs threat modeling (STRIDE/PASTA), code review, and incident response using extensive reference guides for AI agent security, API patterns, and playbooks.

skills · high confidence

Introduces a new Node.js boilerplate for Telegram bots located in skills/telegram/assets/boilerplate/nodejs. The implementation includes a bot client that supports both polling and webhook modes, with webhook mode requiring a 32-256 character secret token for request validation. It features HTML-escaped message handling to prevent injection, automatic retry logic for rate-limit errors, and a health check endpoint. The entry point manages environment-based configuration for token and webhook settings.

skills/telegram/assets/boilerplate/nodejs · high confidence

This change introduces a complete Node.js boilerplate for the WhatsApp Cloud API skill, providing a ready-to-use Express server that handles webhook verification, HMAC signature validation, and message processing. The boilerplate includes a WhatsApp client for sending various message types (text, templates, media, interactive buttons/lists, reactions, locations) with automatic retry logic, a template manager for CRUD operations on WhatsApp templates, and comprehensive TypeScript types for API payloads and webhook events. Users can now integrate WhatsApp messaging capabilities into their Node.js applications using this pre-configured, secure, and type-safe foundation.

skills/whatsapp-cloud-api/assets/boilerplate/nodejs · high confidence

The web application now includes an SVG favicon file in the public directory, replacing or supplementing previous static formats. This ensures the browser tab icon scales cleanly across different display resolutions and device types.

apps/web-app/public · high confidence

The web app now includes \useSkillShortlist\ and \useSkillStars\ hooks to manage user preferences locally in the browser. \useSkillShortlist\ allows users to maintain a persistent list of selected skill IDs, handling storage errors, cross-tab synchronization, and data corruption gracefully. \useSkillStars\ enables users to mark individual skills as saved (starred) with local persistence, supporting both new and legacy storage keys, and preventing duplicate save attempts. These hooks operate independently of server-side state, ensuring that user selections and favorites are retained across sessions even if network or database services are unavailable.

apps/web-app/src/hooks · high confidence

The security-engineer plugin now includes a curated set of offensive and defensive security skills, including Burp Suite testing, cloud penetration testing (Azure, AWS, GCP), ethical hacking methodology, Linux privilege escalation, security auditing, threat modeling, and a reference for the top 100 web vulnerabilities. These skills provide structured workflows, safety gates requiring explicit authorization, and detailed guides for authorized security assessments.

plugins/agentic-bundle-security-engineer/skills · high confidence

The frontend directory for the \todo-app-generated\ example now includes the complete React application source code. This adds the entry point (\main.tsx\), the main application component (\App.tsx\), and styling (\App.css\, \index.css\). It also introduces the core UI components (\TodoForm\, \TodoList\, \TodoItem\, \EmptyState\, \ConfirmDialog\), the custom \useTodos\ hook for state management, and the \api/todos.ts\ module for communicating with the backend. These files constitute the client-side implementation for the example, replacing the previous state where these files were missing or tracked differently.

skills/loki-mode/examples/todo-app-generated/frontend · high confidence

The \docx-official\ skill has been introduced to the \plugins/agentic-bundle-aas-documents-presentations/skills\ directory, providing a new capability for generating Microsoft Word (.docx) documents. This addition includes a detailed tutorial (\docx-js.md\) for using the JavaScript \docx\ library, covering setup, text formatting, styles, lists, and tables, alongside a technical reference (\ooxml.md\) for direct Office Open XML manipulation. The skill also bundles the ISO-IEC 29500-4\_2016 XSD schemas (e.g., \dml-chart.xsd\, \dml-main.xsd\) to support schema validation and compliance, and includes an \LICENSE.txt\ file governing usage under Anthropic's terms.

plugins/agentic-bundle-aas-documents-presentations/skills · high confidence

The web application now includes build scripts to automatically generate and verify critical SEO assets. A new sitemap generator creates an XML file containing the homepage, key navigation routes, SEO landing pages, and a curated list of top skills (defaulting to 180, configurable via environment variables), ensuring only high-signal pages are submitted to search engines. A companion prerender script injects structured data (JSON-LD), meta tags, and canonical links into static HTML pages to improve search discovery. Additionally, a verification script validates these generated assets, checking for correct metadata, secure file paths, and sitemap integrity, with comprehensive test coverage for all helper functions.

apps/web-app/scripts · high confidence

The backend for the todo-app example now includes a complete Express.js API backed by SQLite (via better-sqlite3), featuring CRUD endpoints for todos under /api/todos, automatic database initialization via schema migrations, and a /health check endpoint. Security and stability improvements include rate limiting on the todos routes, input validation for all mutations, graceful shutdown handling, and the removal of the 'x-powered-by' header to reduce information leakage.

skills/loki-mode/examples/todo-app-generated/backend · high confidence

The \docs\_zh-CN\ directory now contains a complete Chinese translation of the project documentation, including the main README, user guides, contributor guidelines, and maintenance notes. A new \.glossary.json\ file (version 1.0.14) provides consistent term translations for 199 key concepts. Many root-level Chinese documentation files have been converted into redirect stubs pointing to their canonical locations in the English documentation structure (e.g., \AUDIT.md\ redirects to \maintainers/audit.md\), ensuring a single source of truth while maintaining Chinese language support for end users.

_docs\zh-CN · high confidence

The skill catalog now automatically organizes skills into meaningful categories (such as Backend, Web Development, AI/ML, and DevOps) instead of leaving them uncategorized. This is driven by a new \generate\_index.py\ script that analyzes skill names, descriptions, and frontmatter to assign categories using keyword matching and dynamic ID-based inference. The generated \skills\_index.json\ now includes \category\_confidence\ and \category\_reason\ fields, allowing users to understand why a skill was placed in a specific category. The web interface reflects this by displaying sorted category filters with skill counts, significantly improving discoverability.

_skill\categorization · high confidence

Adds the \apps/web-app\ directory, a Vite-based React application that serves as a static catalog and skill browser for the Agentic Awesome Skills library. The app renders home, category, and skill detail pages using generated JSON assets, supports local-only star/save interactions, and includes a development-only \refresh-skills\ Vite plugin for updating the catalog. It is configured for GitHub Pages deployment with Tailwind CSS, TypeScript, and optional Supabase integration for read-only skill counts.

apps/web-app · high confidence

Adds the 'Systematic Debugging' skill to the agentic-bundle-essentials, enforcing a four-phase process (Investigation, Pattern Analysis, Hypothesis, Implementation) that mandates root-cause analysis before any code changes. The skill includes a TypeScript implementation for condition-based waiting utilities (replacing arbitrary timeouts with event/state polling) to reduce test flakiness, a Bash/Python diagnostic script to identify test pollution, and documentation on defense-in-depth validation and root-cause tracing. The directory also contains illustrative test scenarios (academic, pressure, and authority tests) to validate the skill's adherence under various constraints.

plugins/agentic-bundle-essentials/skills/systematic-debugging · high confidence

Adds the 'Startup Founder' editorial bundle, providing seven curated skills to help users build products and manage business strategy. The bundle includes specific capabilities for competitive landscape analysis, competitor comparison page creation, conversion-focused copywriting, SaaS launch planning, pricing strategy design, and general product management (featuring RICE prioritization and customer interview analysis scripts). Plugin manifests are added for Claude Code, Codex, and the agent-plugins standard, registering the bundle as version 17.5.0.

plugins/agentic-bundle-startup-founder · high confidence

The \agentic-bundle-aas-product-design-studio/skills\ directory now includes two new agent capabilities: a '3D Web Experience' skill that guides the creation of interactive 3D scenes using Three.js, React Three Fiber, and Spline, and a 'Canvas Design' skill that generates high-fidelity visual artifacts (PDF/PNG) based on abstract design philosophies. These additions are accompanied by the necessary licensing (Apache 2.0) and a bundle of open-source fonts (ArsenalSC, BigShoulders, Boldonse, BricolageGrotesque, CrimsonPro, DMMono, EricaOne) required for the canvas design output.

plugins/agentic-bundle-aas-product-design-studio/skills, plugins/agentic-bundle-creative-director/skills · high confidence

The tools directory now includes a complete v1 implementation for managing agent skills, featuring a new installer (tools/bin/install.js) that supports selective installation for Cursor, Claude Code, Gemini CLI, Codex, Kiro, and Antigravity, along with a new MCP server entry point (tools/bin/aas-mcp.js). The codebase introduces a robust adapter system (tools/lib/aas-v1/adapters) for safely patching host configuration files (Claude's JSON and Codex's TOML) with strict security boundaries, including Windows ACL hardening and symlink protection. Additionally, a new cache layer (tools/lib/aas-v1/cache) handles verified catalog resolution and secure archive extraction, supported by configuration files for audit budgets and generated file tracking.

tools · high confidence

Added a suite of new skills to the \skills\ directory to support AI product development, evaluation, and analytics. These include \ab-test-setup\ for designing and validating A/B tests, \agent-evaluation\ for comparing agent behavior with versioned cases and verifiers, \ai-wrapper-product\ for building focused AI wrapper products, \analytics-product\ for product analytics (PostHog, Mixpanel, cohorts, funnels), \context-window-management\ for LLM context optimization strategies, \hugging-face-evaluation\ for managing evaluation results in Hugging Face model cards, and \kpi-dashboard-design\ for designing effective KPI dashboards. Each skill includes detailed guides, reference materials, and implementation playbooks.

plugins/agentic-bundle-aas-ai-product-evaluation-ops/skills · high confidence

The full-stack developer plugin bundle now includes new instructional skills for API design and backend architecture. The \api-patterns\ skill provides guidance on selecting between REST, GraphQL, and tRPC, along with specific resources for authentication, rate limiting, response formatting, versioning, and security testing (including OWASP API Top 10). It also adds a local heuristic validation script (\api\_validator.py\) for triaging API files. Additionally, the \backend-dev-guidelines\ skill introduces strict architectural doctrines for Node.js/Express/TypeScript services, enforcing a layered architecture (Routes → Controllers → Services → Repositories), mandatory Zod input validation, centralized configuration, and Sentry error tracking, supported by detailed implementation playbooks and examples.

plugins/agentic-bundle-full-stack-developer/skills · high confidence

The \plugins/agentic-bundle-aas-api-platform-builder/skills\ directory now includes a comprehensive set of new skills for API development. These include \api-design-principles\ (with checklists, templates, and reference guides for REST and GraphQL), \api-documentation\ (a workflow for OpenAPI specs and developer guides), \api-endpoint-builder\ (for generating production-ready endpoints with validation and security), and \api-patterns\ (covering style selection, auth, rate limiting, and security testing). These additions provide structured guidance and templates for designing, documenting, and building APIs within the platform.

plugins/agentic-bundle-aas-api-platform-builder/skills, plugins/agentic-bundle-aas-qa-test-automation/skills · high confidence

This update introduces three new agentic skills to the toolkit. The Analytics Tracking skill provides a structured framework and diagnostic index for auditing and improving analytics instrumentation to ensure data reliability. The Apify Market Research skill enables automated extraction of market, pricing, and consumer behavior data from platforms like Google Maps, Facebook, and TripAdvisor using Apify Actors. The Content Creator skill offers a complete workflow for drafting, optimizing, and planning content, including local Python scripts for brand voice analysis and SEO diagnostics, along with templates for content calendars and brand guidelines.

plugins/agentic-bundle-aas-localization-international-growth/skills · high confidence

A new 'app-store-optimization' skill has been added to the skills bundle, providing a comprehensive toolkit for researching, optimizing, and tracking mobile app performance on the Apple App Store and Google Play Store. This skill includes eight Python modules for keyword analysis, metadata optimization, competitor intelligence, ASO health scoring, A/B testing planning, localization, review analysis, and launch checklists, along with documentation and example inputs to guide users through tasks like keyword research, metadata optimization, and competitive gap analysis.

plugins/agentic-bundle-expo-react-native/skills · high confidence

A new 'app-store-optimization' skill has been added to the skills directory, providing a comprehensive toolkit for optimizing mobile app presence on the Apple App Store and Google Play Store. This skill includes Python modules for keyword analysis, metadata optimization, competitor intelligence, ASO health scoring, A/B testing planning, localization, review analysis, and launch checklists. It enables users to research keywords, optimize titles and descriptions with platform-specific character limits, analyze competitors, calculate ASO scores, plan statistical A/B tests for icons and screenshots, manage multi-language localization, analyze user review sentiment, and generate pre-launch compliance checklists.

plugins/agentic-bundle-aas-mobile-app-builder/skills · high confidence

A new 'app-store-optimization' skill has been added to the mobile developer bundle, providing a comprehensive toolkit for researching, optimizing, and tracking mobile app performance on the Apple App Store and Google Play Store. This skill includes eight Python modules for keyword analysis, metadata optimization, competitor intelligence, ASO health scoring, A/B testing planning, localization, review analysis, and launch checklists, along with documentation and example outputs to guide users through improving their app's visibility and conversion rates.

plugins/agentic-bundle-mobile-developer/skills · high confidence

A new 'docx-official' skill has been added to the documents and presentations bundle, providing structured guidance for creating, editing, and analyzing .docx files. The skill introduces specific workflows for generating documents via the docx-js library, performing OOXML manipulation with a Python Document library, and executing complex tracked-change (redlining) operations. It also includes a full reference of ISO-IEC 29500-4 XSD schemas and technical documentation to support these document operations.

plugins/agentic-bundle-documents-presentations/skills · high confidence

A new 'git-pushing' skill has been added to the agentic bundle, providing a safe, atomic workflow for staging, committing, and pushing changes. The skill introduces a \smart\_commit.sh\ helper that enforces conventional commit messages, prevents mixed-index pushes by using an isolated temporary index, and ensures branch safety through file locking and concurrent-change detection. It respects repository policies, such as protected branches and remote configurations, and rejects operations on incompatible Git ref backends (e.g., reftable). Users can now explicitly request to push changes, and the system will handle the staging and commit process securely, ensuring that only intended changes are committed and pushed.

plugins/agentic-bundle-essentials/skills/git-pushing · high confidence

The \plugins/agentic-bundle-aas-python-api-builder/skills\ directory now includes a suite of new skills to guide the development of Python-based APIs and asynchronous applications. These additions provide structured guidance and implementation playbooks for API design principles (covering REST and GraphQL best practices, checklists, and templates), FastAPI project setup and advanced patterns, Django 5.x architecture, and asynchronous Python concurrency patterns using asyncio. Users can now leverage these skills to generate production-ready code structures, enforce design standards, and implement robust error handling and security measures in their Python backend services.

plugins/agentic-bundle-aas-python-api-builder/skills, plugins/agentic-bundle-business-analyst/skills · high confidence

Added a new collection of agent skills in the \plugins/agentic-bundle-qa-testing/skills\ directory to standardize quality assurance and debugging workflows. The bundle includes \browser-automation\ for reliable UI testing with semantic locators, \ci-cd-and-automation\ for enforcing quality gates in pipelines, \code-review-checklist\ for systematic code audits, \e2e-testing-patterns\ for building stable end-to-end test suites, and \systematic-debugging\ for root-cause analysis and condition-based waiting. These skills provide structured guidance, examples, and reference materials to help agents perform consistent, high-quality testing and debugging tasks.

plugins/agentic-bundle-qa-testing/skills · high confidence

Introduces the 'SEO Specialist' bundle, a collection of seven specialized skills for technical SEO, content structure, and search growth. The bundle includes skills for schema markup design and validation, SEO cannibalization detection, content auditing (E-E-A-T), content planning and outlining, SEO-optimized content writing, core SEO fundamentals, and content structure architecture. It also provides a Python-based SEO checker script to audit HTML/JSX/TSX pages for issues like missing meta tags, heading hierarchy errors, and missing alt attributes. The bundle is packaged for integration with Claude Code, Codex, and the agent-plugins ecosystem.

plugins/agentic-bundle-seo-specialist · high confidence

The 'Skill Author' bundle is now available, providing a curated set of six skills to help users author, review, and maintain high-quality agent SKILL.md assets. This release includes the 'Skill Creator' for automating the brainstorming, template application, and validation of new CLI skills, 'Effective Agent Skills' for guidance on triggers and safety, 'Lint and Validate' with helper scripts for running project linters and checking type coverage, and 'Documentation Generation' for creating API and architecture docs from code. The bundle is distributed via plugin manifests for Claude Code, Codex, and the agent-plugins standard.

plugins/agentic-bundle-skill-author · high confidence

The 'TypeScript & JavaScript' bundle from Agentic Awesome Skills is now available, providing five curated skills to guide AI agents in modern web development. The bundle includes 'JavaScript Pro' for ES6+ and async patterns, 'Next.js App Router Patterns' for Server Components and full-stack React, 'Node.js Best Practices' for architecture and framework selection, and 'React Best Practices'—a comprehensive Vercel-maintained guide covering performance optimization, bundle size reduction, and waterfall elimination. This update adds the necessary plugin manifests and skill definitions to enable these capabilities in supported agent environments.

plugins/agentic-bundle-typescript-javascript · high confidence

The \plugins/agentic-bundle-aas-accessibility-inclusive-ux/skills\ directory now includes a suite of new agent skills for accessibility auditing and browser automation. These include \accesslint-audit\ for finding and fixing WCAG 2.2 issues via live-DOM or static audits, \accesslint-diff\ for comparing accessibility violations against a baseline, and \accesslint-scan\ for reporting violations without editing. Additionally, \playwright-skill\ provides a universal executor for writing and running custom Playwright automation scripts, while \fixing-accessibility\ and \ui-a11y\ offer guidelines and static audits for HTML accessibility and design-system token compliance. The bundle also adds \screen-reader-testing\ with a detailed implementation playbook for validating assistive technology compatibility, and \webapp-testing\ with its licensing information.

plugins/agentic-bundle-aas-accessibility-inclusive-ux/skills, plugins/agentic-bundle-architecture-design/skills, plugins/agentic-bundle-indie-game-dev/skills · high confidence

The \plugins/agentic-bundle-agent-architect/skills\ directory now includes a curated set of documentation-based skills for building and managing AI agents. This adds the \agent-evaluation\ skill for running regression tests and measuring reliability, the \agent-memory-systems\ skill for implementing short-term and long-term memory architectures, the \ai-agents-architect\ skill for designing autonomous agent loops and safety guardrails, the \langgraph\ skill for constructing stateful multi-actor graphs, and the \mcp-builder\ skill for creating Model Context Protocol servers. These files provide the procedural guides, reference patterns, and evaluation fixtures that agents use to perform these tasks.

plugins/agentic-bundle-agent-architect/skills · high confidence

The skills directory now includes new documentation for five capabilities: Airtable, GitHub, Google Sheets, and Make automation via Rube MCP (Composio), plus an MCP server development guide. These files define the prerequisites, core workflows, tool sequences, and pitfalls for automating records and bases in Airtable, managing GitHub issues and pull requests with repository-policy safeguards, reading and writing Google Sheets data, and retrieving Make operation logs and enums. The bundle also adds the Apache 2.0 license for the mcp-builder skill and reference guides for MCP best practices and bounded evaluation.

plugins/agentic-bundle-aas-automation-builder/skills, plugins/agentic-bundle-automation-builder/skills, plugins/agentic-bundle-marketing-growth/skills · high confidence

The web app now includes dedicated pages for specialized plugins, topic-based landing guides, and a workbench for reviewing agent-selected skill stacks. The Home page has been redesigned with a virtualized grid, advanced filtering (category, risk, source, scope), and URL-synced search state for shareable bookmarks. Skill detail pages now support lazy-loaded markdown rendering with security checks against HTML injection, and link resolution is anchored to the pinned catalog release version. A new 404 page provides clear navigation back to the catalog. Comprehensive test coverage has been added for the Home, Plugins, and SkillDetail pages to ensure stability of these new features.

apps/web-app/src/pages · high confidence

Added a suite of new agent skills in the \plugins/agentic-bundle-aas-data-analytics/skills\ directory to support data-driven decision-making and visualization. These include \ab-test-setup\ for designing and validating A/B tests with strict hypothesis and tracking gates; \analytics-product\ for product analytics workflows using tools like PostHog and Mixpanel, including cohort and funnel analysis; \analytics-tracking\ for designing and auditing event tracking systems with a measurement readiness index; \business-analyst\ for strategic KPI frameworks, financial modeling, and BI tool usage; and \claude-d3js-skill\ for creating custom, interactive data visualizations using D3.js, complete with code templates, color scheme references, and sample data.

plugins/agentic-bundle-aas-data-analytics/skills · high confidence

The lint-and-validate skill is now available to run configured lint and type checks for Node.js and Python projects, distinguishing actual failures from checks that could not run. It uses a new lint\_runner.py script that detects project types and executes safe, read-only commands (e.g., npm run lint, tsc --noEmit, ruff check, mypy) without installing dependencies or applying fixes, reporting exact results and any missing tools. A new type\_coverage.py script provides a bounded, read-only annotation inventory (up to 30 files per language) that counts fully annotated Python functions and explicit TypeScript any occurrences, explicitly noting that this is not a type-safety verdict and should not replace the project's configured type checker.

plugins/agentic-bundle-essentials/skills/lint-and-validate · high confidence

Added new agent skills for A/B test setup, analytics tracking, content creation, copywriting, and email sequences. These skills provide structured workflows and local analysis tools (Python scripts for brand voice and SEO optimization) to help users design experiments, audit data quality, draft audience-specific content, write conversion-focused copy, and plan email nurture campaigns.

plugins/agentic-bundle-aas-marketing-seo-growth/skills · high confidence

The Agentic Awesome Skills bundle now includes a comprehensive set of new security testing and auditing capabilities. Users can perform API security testing (covering REST/GraphQL authentication, authorization, and input validation), execute web application assessments using Burp Suite, and conduct cloud penetration testing across Azure, AWS, and GCP with specific scripts for enumeration and exploitation. The bundle also adds ethical hacking methodology guides, Linux privilege escalation workflows, and Static Application Security Testing (SAST) configuration instructions for tools like Semgrep and SonarQube. Additionally, a new security auditor skill provides guidance on DevSecOps, compliance frameworks (OWASP, GDPR, PCI-DSS), and secure coding practices. All offensive skills include mandatory authorization confirmation gates to ensure responsible use.

plugins/agentic-bundle-aas-security-engineer/skills · high confidence

Added \activate-skills\ (Windows batch and Unix shell) to manage the Antigravity skills library, including syncing from the repository, merging from archives, expanding bundles via a Python helper, and restoring selected skills with safe ID validation. Added \validate-links.sh\ to scan Markdown files in the documentation directories for broken internal links and report them. Added \validate-glossary.sh\ to check the Chinese glossary JSON for structural validity, term count consistency, and missing translations.

scripts · high confidence

The \plugins/agentic-awesome-skills/skills\ directory now includes two new agent skills: \00-andruia-consultant\, which acts as a solutions architect to diagnose project status and define technical roadmaps in Spanish, and \007\, a comprehensive security audit and hardening skill that provides threat modeling (STRIDE/PASTA), OWASP checklists, incident response playbooks, and reference guides for API and AI agent security. These additions expand the plugin's capabilities to cover initial project architecture and rigorous security validation.

plugins/agentic-awesome-skills/skills · high confidence

The plugin now includes a new '007' security skill that provides comprehensive security audit, hardening, and threat modeling capabilities (using STRIDE/PASTA frameworks) for general projects, alongside specialized reference guides for AI agent and LLM pipeline security. It also adds an 'Andru.ia Solutions Architect' skill to help diagnose project status and create technical roadmaps in Spanish. These new skills introduce new operational modes for security analysis, including incident response playbooks and OWASP checklists, expanding the plugin's ability to assist with security compliance and architectural planning.

plugins/agentic-awesome-skills-claude/skills · high confidence

The web app now includes a set of new components to help users find, compare, and install skills. The OutcomeExplorer component allows users to search for skills by describing their goal, viewing matching results with evidence and gaps, and adding candidates to a browser-local shortlist. The ShortlistReview component provides a comparison table for selected skills, showing their requirements and provenance, and allows users to generate and copy a focused agent brief tailored to a specific target (Codex or Claude). The InstallationHandoff component generates a dry-run installation preview command based on the selected skills and user-specified directory, while the SelectionFeedback component lets users record local, voluntary feedback about their selection experience. Additionally, the SkillCard component has been updated to display risk levels and include a separate save button outside the main link, and a new Icon component centralizes UI icons.

apps/web-app/src/components · high confidence

The web application now includes dedicated data structures to support targeted SEO landing pages and specialized plugin discovery. A new \seoLandingPages.json\ file defines metadata (titles, descriptions, keywords, and featured skill IDs) for specific topic hubs such as Antigravity CLI skills and the GitHub AI skills repository. The accompanying \seoLandingPages.ts\ module provides logic to retrieve these pages and curate a deterministic list of relevant skills for each page based on editorial picks and topic scoring. Additionally, \specializedPlugins.ts\ loads and exposes specialized plugin definitions by merging candidate manifests with editorial bundle data, enabling the UI to display focused plugin packs for domains like security, DevOps, and data analytics.

apps/web-app/src/data · high confidence

The \skills/vercel-optimize\ module has been significantly expanded with new library files that harden input validation and expand analysis capabilities. A new \auth-route.mjs\ module prevents caching of sensitive user-state routes (login, account, checkout, etc.) at the CDN edge by disqualifying them from cache optimization. A \citations.mjs\ module curates documentation references, filtering them by framework and version to ensure recommendations are relevant and safe. A \cost-coverage.mjs\ module maps billing line items to audit dimensions, surfacing uncovered costs (like Sandbox or AI Gateway) as blind spots. Additionally, new modules for budget summaries, recommendation deduplication, deep-dive query specs, display labeling, claim extraction, and framework support gates provide a more robust, verifiable, and user-friendly optimization workflow.

skills/vercel-optimize · high confidence

The 'Web Designer' plugin bundle (v17.5.0) now includes three new curated skills: a 3D Web Experience skill for building interactive 3D scenes using Three.js, React Three Fiber, and Spline; an Accessibility Audit skill for conducting WCAG compliance checks and remediation; and a Canvas Design skill for generating high-fidelity visual philosophies and artwork. The bundle also adds bundled open-source fonts (ArsenalSC, Big Shoulders, Boldonse, BricolageGrotesque) to support the canvas design output.

plugins/agentic-bundle-web-designer · high confidence

A new migration (202607300001) enforces Row Level Security on the public.skill\_stars table, restricting the anon and authenticated roles to SELECT-only access and revoking all write privileges (POST, PATCH, DELETE). The migration also revokes write access on all public sequences to prevent potential privilege escalation through sequence manipulation, addressing coordinated security findings.

supabase · high confidence

The web application has been restyled with a new visual identity, including a custom logo and the 'AAS Core' brand name in the header. The navigation structure has been updated to feature 'Core', 'Workbench', and 'Plugins' links, with a mobile menu and GitHub integration. The layout now uses a sticky header with a blurred background and supports both light and dark modes via CSS variables. Page content is wrapped in a suspense boundary for lazy-loaded routes, improving perceived performance.

apps/web-app/src · high confidence

The project has been renamed from its previous title to 'Agentic Awesome Skills' (AAS Core). This change updates the repository's public identity and narrative, aligning the README, documentation, and release notes with the new branding to clarify the scope of the curated skill collection.

(repo-wide) · high confidence

The AAS v1 skill catalog has been updated to package version 17.5.0, now containing 2,127 skills. This release includes new skill entries such as \007\ (AI agent security), \2slides-ppt-generator\, and \agent-orchestrator\, alongside updates to the underlying \catalog-manifest.v1.json\ and \skill-content-index.v1.json\ files to reflect the new content and schema version 2.0.0.

data/aas-v1 · high confidence

The web app now uses a new SkillContext to load the skills catalog with improved reliability. It attempts to fetch the skills index from multiple candidate URLs to handle different deployment paths (such as GitHub Pages), validates that the response is a proper JSON array, and ignores stale or failed requests. Star counts are fetched asynchronously so they do not block the main skills data, ensuring the catalog remains available even if the star service is slow or unavailable.

apps/web-app/src/context · high confidence

Add comprehensive test coverage for web-app utility functions; Added test environment setup with browser API mocks; Added tests for SkillContext loading and validation logic; Added tests for route lazy loading, Workbench isolation, and refresh-skills security; New AAS agent-first preview verification suite.

The repository now includes \package.json\ and \package-lock.json\ files for the root \agentic-awesome-skills\ package and the \apps/web-app\ sub-application, establishing the dependency graph for the project. The web app specifies React 19.2.8, Vite 8.0.16, and Tailwind CSS 4.2.0, while the root package bundles \ajv\, \fast-uri\, \sanitize-filename\, and \yaml\ for CLI tooling. Additionally, \requirements.txt\ files are added to various skill scripts to declare Python dependencies such as \fastapi\, \httpx\, and \google-genai\.

(dependencies) · high confidence

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

• First survey — no prior run to compare against. CAI 47.

• Code Health 41
• Architecture 83
• Maturity 76
• Readiness 48
• Security 74
• Accessibility 40

• 300 commits — 250 feature/other, 50 fixes

• (root) — 86 commits
• apps/web-app — 34 commits
• docs/users — 31 commits
• data/aas-v1 — 26 commits
• plugins/agentic-awesome-skills — 20 commits
• tools/scripts — 15 commits
• (repo) — 5 commits
• skills/loki-mode — 5 commits
• plugins/agentic-awesome-skills-claude — 4 commits
• skills/ui-slop-score — 3 commits
• .github/workflows — 2 commits
• docs/maintainers — 2 commits
• skills/api-rate-limit-handler — 2 commits
• skills/falsify — 2 commits
• skills/graceful-shutdown — 2 commits
• skills/multi-source-search — 2 commits
• skills/project-state-governor — 2 commits
• skills/unified-ai-gateway — 2 commits
• docs/README.md — 1 commit
• docs/examples — 1 commit

• fix: fix(skill-content): entropy-box security/license/Limitations/CRLF + README (#1331)
• fix: fix(web): ignore stale catalog refresh results (#1459)
• fix: fix(web): keep shortlist aligned with persistence (#1448)
• fix: fix(web): only mark persisted skills as saved (#1446)
• fix: fix(web): validate catalog records before use (#1464)
• fix: fix(workbench): enforce published profile limits (#1465)
• fix: fix(workbench): honor plan identity schema (#1489)
• fix: fix: admit explicitly reviewed fork skill bundles (#1415)
• fix: fix: align project state lifecycle statuses (#1241)
• fix: fix: allow zero-diff contributor merges (#1171)
• fix: fix: allow zero-diff fork preflight (#1173)
• fix: fix: bind installer clones to npm release identity (#1132)
• fix: fix: classify copies into existing skills (#1362)
• fix: fix: classify zero-diff contributor PRs (#1172)
• fix: fix: complete reviewed maintenance repairs (#1244)
• fix: fix: complete specialized plugin resources and executable examples (#1380)
• fix: fix: correct Tree Ring skill guidance (#1253)
• fix: fix: exclude generated Python cache from npm packages (#1352)
• fix: fix: harden archive and installer boundaries and browser reviews (#1376)
• fix: fix: harden multi-source evidence integrity (#1205)
• …and 280 more

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

sickn33/agentic-awesome-skills 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.