Yaz-inc

skill-template

Template only — copy this folder and rename. Do not invoke directly.

ai-evaluation-red-team

Design and execute repeatable evaluations and authorized red-team tests for prompts, models, agents, tools, and RAG systems across quality, safety, security, reliability, cost, and latency. Use before releasing or materially changing AI behavior.

ai-observability

Design privacy-aware observability for AI applications across prompts, models, retrieval, agents, tools, quality, latency, cost, tokens, errors, traces, feedback, and evaluation results. Use for production readiness and AI incident diagnosis.

analyst-insight-reporting

Perform reproducible exploratory and decision-focused analysis, calculate business metrics, test assumptions, create accessible visualizations, and communicate findings with uncertainty and provenance. Use for analyst reports, dashboards, investigations, and executive insights.

analytics-engineering

Design tested analytics transformations, staging layers, dimensions, facts, semantic metrics, documentation, freshness controls, and reproducible builds. Use when turning operational data into trusted reporting and decision datasets.

animated-glow-border

Add a polished, theme-aware rotating glow around cards, panels, certificates, or feature containers using CSS conic gradients. Use for animated borders, traveling light effects, reverse border motion, glowing card outlines, live UI demonstrations, or subtle emphasis on dark and light interfaces.

Testing & QA Listed

api-contract-testing

Validate REST and HTTP APIs against OpenAPI contracts, authorization rules, data invariants, malformed inputs, error responses, pagination, concurrency, and backward compatibility. Use for API audits, automated contract tests, and pre-release verification.

application-security-review

Perform evidence-based security reviews of web applications and APIs using risk-appropriate threat modeling, OWASP ASVS and API guidance, static analysis, authorization testing, and remediation verification. Use before release or after material security changes.

big-data-analytics

Select and operate scalable analytics workflows for datasets that exceed practical single-process memory or runtime limits using distributed or parallel computation. Use for large batch, streaming, iterative, or partitioned analysis.

central-knowledge-base

Maintain and retrieve shared, cross-project knowledge in a central local directory without placing secrets or confidential client data in Markdown. Use when the user asks to add information to the central KB, create a shared source of truth, document reusable project references, reconcile duplicate knowledge, or help another agent discover approved resources.

DevOps & Infrastructure Listed

ci-cd-release-gates

Design and review continuous integration and delivery pipelines with build, test, security, artifact, approval, deployment, observability, and rollback gates. Use when creating or hardening release automation across application stacks.

Data & Documents Listed

data-lineage-governance

Document and operate data lineage, cataloging, ownership, classifications, retention, quality, access, impact analysis, and governance for data and AI systems. Use when datasets have multiple producers, transformations, consumers, or regulatory and contractual obligations.

Data & Documents Listed

data-pipeline-orchestration

Design, implement, and review resilient data pipelines with scheduling, dependencies, idempotency, retries, backfills, checkpoints, quality gates, observability, ownership, and recovery. Use for ETL, ELT, batch, event, and analytics workflows.

Data & Documents Listed

data-profile-clean-validate

Profile, clean, normalize, and validate structured or semi-structured datasets with reproducible transformations, schema checks, anomaly evidence, and before-and-after reconciliation. Use for CSV, spreadsheet, Parquet, JSON, dataframe, and extracted database data.

data-quality-contracts

Define and enforce versioned data contracts for schemas, keys, nullability, domains, freshness, volume, ownership, compatibility, quality thresholds, and incident response. Use between data producers and consumers or before publishing analytical data.

data-reconciliation-migration

Plan and verify data migrations, imports, exports, corrections, and system synchronization using counts, keys, hashes, aggregates, referential checks, exception reports, cutover controls, and rollback. Use whenever data moves or changes at scale.

database-reliability-audit

Audit relational database integrity, constraints, transactions, concurrency, indexes, backups, restores, retention, observability, and recovery readiness. Use for production-readiness reviews, migration planning, data-risk assessments, or recurring database health checks.

deepeval-ai-testing

Create maintainable AI regression tests with DeepEval, versioned datasets, task-appropriate metrics, pytest integration, traces, thresholds, and failure analysis. Use when an AI application needs repeatable local or CI evaluation.

deploy-secrets-zero-exposure

Strict pre/post-deploy gate: no secrets, keys, .env, .git, ops files, or credential artifacts reachable on the public domain or to unauthorized users. Use before every deploy, when adding fix.php or admin tools, reviewing hosting, or when the user asks about deployment security or secret exposure online.

Web & Frontend Listed

design-system-engineering

Create, audit, or synchronize design tokens, reusable components, variants, documentation, and code across design tools and product repositories. Use for Figma libraries, frontend component libraries, typography and color systems, responsive foundations, theming, governance, design-code parity, or reducing inconsistent UI patterns.

Web & Frontend Listed

design-to-code-implementation

Implement approved Figma, Canva-derived assets, screenshots, prototypes, or written UI specifications as maintainable production interfaces. Use for design handoff, responsive frontend work, component mapping, visual fidelity, interaction states, accessibility, browser verification, or closing gaps between a design and its running implementation.

figma-canva-design-workflow

Coordinate Figma product-interface design with Canva campaign, presentation, social, and lightweight brand production. Use when a project needs assets across both platforms, a controlled design handoff, shared brand consistency, export specifications, licensing checks, or implementation-ready UI files alongside marketing collateral.

laravel-activity-logger

Implements hybrid activity audit logging in Laravel — indexed columns for fast queries plus JSON payload for full context. Use when adding audit trails, activity log pages, or logging create/update/delete across admin modules.

laravel-ai-multi-provider

Multi-provider AI chat and vision with automatic fallback when quota, rate limits, or model unavailability occur. Use when integrating Gemini, OpenRouter, OpenAI, or Claude in Laravel with resilient error messages.

laravel-datatables-module

Builds server-side Yajra DataTables CRUD list modules in Laravel admin apps. Use when adding index pages with AJAX tables, filters, action columns, or migrating modal CRUD to separate index/create/edit pages.

laravel-deploy-fix-php

Post-deploy Laravel cache recovery via public/fix.php without SSH or exec(). Use after git pull when app shows 500 from stale config/route cache, or when setting up production deploy recovery on shared hosting.

laravel-dual-repo-push

Push one tested branch or release tag to separate developer and client GitHub remotes using clean remote URLs and an approved credential manager. Use when a Laravel or other software project must deliver the same verified commit to two repositories without embedding tokens in commands, scripts, configuration, or logs.

laravel-env-maintenance

Toggles Laravel maintenance mode from .env (APP_MAINTENANCE) with branded 503 page, secret bypass URL, and safe coexistence with manual artisan down. Use when deploying upgrades without SSH or for scheduled maintenance windows.

laravel-form-select2-global

App-wide consistent height and padding for Bootstrap form-control, form-select, and Select2 fields via CSS variables in the main layout stylesheet. Use when Select2 dropdowns look misaligned on some pages but not others in Bootstrap 5 admin UIs.

laravel-json-ai-settings

Stores AI provider settings and encrypted API keys in JSON under storage, editable via admin UI without .env changes. Use when building Laravel apps with configurable AI keys, provider selection, or runtime AI settings.

laravel-rbac-permissions

Implements module-based RBAC in Laravel admin apps with permission middleware, user_permissions table, super_admin bypass, and route groups. Use when adding role-based access, gating routes by module read/write, or staff vs portal users.

laravel-security-hardening

Production security checklist for Laravel admin apps — lock database setup, gitignore ops files, validation rule against XSS payloads, and safe 403/404 responses. Use when hardening deploys or reviewing production exposure.

laravel-soft-delete-trash

Unified soft-delete trash dashboard with deleted_by tracking, 30-day expiry, restore, and scheduled force-delete. Use when adding SoftDeletes across modules, recycle bin UI, or trash:cleanup artisan command.

laravel-vision-label-scan

Vision AI pipeline for Laravel — upload label photo, extract JSON fields, user reviews and applies to form (never auto-save). Use for OCR/VLM label scanning on admin forms, asset registration, or catalog model entry.

mission-control

Interact with Mission Control — AI agent orchestration dashboard. Use when registering agents, managing tasks, syncing skills, or querying agent/task status via MC APIs, OpenClaw adapters, or SSE events.

nosql-query-engineering

Design and validate data models and complex queries for document, search, key-value, cache, and related NoSQL systems, including aggregation, indexing, consistency, partitioning, authorization, and operational tradeoffs. Use when relational assumptions do not fit the selected store.

Testing & QA Listed

playwright-e2e-testing

Plan, implement, and verify cross-browser end-to-end tests with Playwright for authenticated workflows, CRUD journeys, responsive behavior, failure evidence, and regression coverage. Use when an application needs repeatable browser automation rather than one-time manual testing.

rag-evaluation

Evaluate retrieval-augmented generation systems across ingestion, chunking, indexing, retrieval, reranking, context precision, context recall, faithfulness, answer relevance, citations, freshness, and tenant isolation. Use for RAG development and release gates.

Testing & QA Listed

screen-reader-accessibility-testing

Test websites and applications through screen readers, keyboard-only navigation, semantic inspection, and assistive-technology behavior. Use for NVDA, JAWS, VoiceOver, TalkBack, Narrator, machine-reader compatibility, accessible names and roles, live regions, focus management, forms, dialogs, data tables, dynamic updates, or WCAG 2.2 release evidence.

sql-query-engineering

Design, explain, validate, and safely execute complex SQL involving joins, CTEs, window functions, recursive logic, aggregates, pivots, dates, reconciliation, and controlled writes across SQL dialects. Use for analysis, reporting, data fixes, or backend query development.

test-software-app

Profile, plan, execute, remediate, and report human-centered software application testing and pre-go-live audits. Use when Codex must test a web app, portal, API-backed product, internal system, SaaS product, mobile/PWA experience, or desktop application across CRUD data behavior, roles and tenant isolation, search, sort, filters, exports, email, chat, reports, integrations, security, regression, and deployment readiness. Supports selectable 3-stage, 4-stage, 5-stage, and 6-stage audit depth, including audit-only engagements, authorized fixes, database exclusions, QA record cleanup, and evidence-backed client certificates.

accessibility-audit

Audit web accessibility using automated checks and human keyboard, focus, zoom, reflow, semantic, form, contrast, authentication, and assistive-technology workflows. Use for WCAG 2.2 Level AA quality targets and release verification.

architecture-decision-records

Create and maintain concise architecture decision records that capture context, options, decision drivers, consequences, security, operations, reversibility, and follow-up evidence. Use for material technical choices and changes with lasting cost.

data-operations-router

Classify data requests by objective, source, scale, sensitivity, mutation risk, and evidence needs, then route work to the correct specialized toolkit skill. Use before cleaning, querying, migrating, validating, analyzing, or operating unfamiliar data.

extract-to-toolkit

Extracts reusable patterns from a finished or active project into Yazinc Ai Toolkit. Use when the user asks to add skills to the toolkit, publish patterns, extract from a project, close a project, or push learnings to yazinc-ai-toolkit. Reads AGENTS.md and WORKFLOW.md in the toolkit repo first.

frontend-quality-gate

Define and enforce frontend quality budgets for performance, responsive behavior, SEO, best practices, console errors, network failures, and regression evidence. Use for web release gates and recurring quality monitoring.

observability-incident-response

Design and assess application observability and incident response across logs, metrics, traces, correlation, alerting, runbooks, recovery, and post-incident learning. Use for production readiness, reliability improvement, or active incident diagnosis.

performance-load-testing

Plan and execute load, stress, spike, soak, capacity, and API performance tests with controlled traffic, service-level thresholds, bottleneck evidence, and safe release decisions. Use when response time, throughput, concurrency, or scalability matters.