smartrus
31 verified skills
agentic-rag-evaluator
Acts as the verification and grounding layer in Retrieval-Augmented Generation (RAG) pipelines. Evaluates retrieved context chunks for factual accuracy, relevance, and completeness before they reach the generation stage. Detects hallucination risk, stale data, source conflicts, and insufficient context. Trigger when: rag evaluation, rag quality, retrieval quality check, grounding check, hallucination detection, context verification, rag pipeline audit, retrieval accuracy, evaluate rag. Do NOT trigger: building RAG pipelines from scratch, vector database setup or tuning, prompt engineering for LLMs, general data quality auditing without RAG context, or training data validation (use data-poisoning-auditor).
development
ai-citation-strategist
Focuses on Answer Engine Optimization (AEO) to maximize content visibility and citation frequency across generative AI systems (ChatGPT, Perplexity, Claude, Google AI Overviews). Analyzes content structure, authority signals, and schema markup to ensure content is cited by AI answers rather than just ranked by traditional search. Do NOT trigger on traditional SEO keyword research only, general content writing without AI citation focus, social media marketing, paid advertising, or email marketing.
development
aiops-telemetry-sanitizer
Cleanses observability logs and telemetry streams to defend autonomous AIOps agents against adversarial prompt injections and reward-hacking embedded in log data. Scans log lines for injection patterns, suspicious Unicode, and encoded payloads that could manipulate LLM-based alerting or automated remediation systems. Trigger when: log sanitization is needed, telemetry cleansing required, prompt injection suspected in logs, aiops security hardening, log injection defense implementation, adversarial log detection, reward hacking defense needed, or when you need to sanitize telemetry pipelines. Do NOT trigger for: general log aggregation or search, application-level logging configuration, setting up monitoring dashboards, Kubernetes pod debugging, SIEM rule writing, or general security scanning.
development
api-version-migrator
Safely adapts codebases across breaking library changes by understanding nuances between older and newer API versions. Analyzes import statements, API call signatures, and dependencies to identify deprecated features and create step-by-step migration plans. Helps teams upgrade libraries, SDKs, and frameworks while maintaining code functionality. Do NOT trigger: general code refactoring without version context, writing new APIs from scratch, API design or documentation, performance optimization
development
autonomous-rollback-orchestrator
Triggers automated remediation and rollback pipelines when real-time health metrics breach Service Level Objectives (SLOs). Designs rollback decision trees, configures metric-based triggers, and generates runbooks for autonomous incident response that minimize MTTR while respecting human-in-the-loop checkpoints. Trigger when: automated rollback needed, SLO breach detected, rollback orchestration required, autonomous remediation needed, metric-based rollback decisions, rollback pipeline design, rollback decision trees, or canary rollback scenarios. Do NOT trigger for: manual incident response workflows, general CI/CD pipeline setup, Terraform drift detection, application performance tuning, or monitoring/alerting configuration.
testing
carousel-growth-engine
Autonomously designs, formats, and optimizes multi-slide carousel content specifically engineered for social media algorithm engagement on LinkedIn, Instagram, and Twitter/X. Applies platform-specific best practices for slide count, text density, visual hooks, and CTA placement to maximize impressions and shares. Do NOT trigger on general social media posting or scheduling, presentation/PowerPoint creation, infographic design without carousel context, blog post writing, or email newsletter design.
development
cloud-finops-optimizer
Infrastructure cost right-sizer and FinOps optimization specialist. Analyzes cloud resource utilization, identifies overprovisioned or idle resources, recommends rightsizing actions, generates cost-saving reports, and produces Terraform/IaC patches for implementing changes. Supports AWS, GCP, and Azure. Use this skill whenever the user mentions cloud costs, cloud spend, rightsizing, reserved instances, savings plans, overprovisioned resources, idle resources, cost optimization, FinOps, cloud billing, compute waste, or cost anomalies — even if they don't explicitly say "FinOps". Do NOT trigger when the user is asking about application performance tuning without cost context, Kubernetes pod scheduling or debugging (use k8s-debugger instead), infrastructure provisioning from scratch (use IaC tools directly), or security compliance auditing.
tools
cross-cultural-ux-auditor
Analyzes UI layouts, marketing copy, imagery, iconography, and color usage to detect cultural bias, misrepresentation, and localization failures that could alienate global audiences. Provides culturally-aware recommendations grounded in Hofstede dimensions, color psychology by culture, and regional UX conventions. Do NOT trigger on general UX design review without cultural focus, language translation only, accessibility/WCAG auditing, general design critique, or market research.
testing
data-poisoning-auditor
Scans AI training data pipelines and datasets to detect bias, data poisoning, label corruption, and statistical anomalies that could compromise autonomous AI decision-making. Generates audit reports with poisoning risk scores and remediation recommendations. Trigger when: data poisoning, training data audit, dataset bias, label corruption, data integrity check, poisoned data detection, training pipeline audit, adversarial data, dataset contamination. Do NOT trigger: general data quality profiling without AI/ML context, RAG evaluation (use agentic-rag-evaluator), model performance evaluation, feature engineering, or data visualization.
testing
digital-twin-dashboard-designer
Engineers the frontend representations and data visualization layers for digital twin systems, making complex real-time telemetry, sensor data, and simulation outputs accessible through intuitive dashboards. Designs alert hierarchies, anomaly highlighting, and drill-down interaction patterns for operational monitoring. Trigger on digital twin dashboards, telemetry visualization, real-time monitoring UI, and operational dashboards. Do NOT trigger on digital twin specification/modeling, general web dashboard design, data analytics dashboards, backend data pipelines, or general data visualization without real-time operational context.
development
digital-twin-planner
Models virtual replicas of physical systems (factories, supply chains, infrastructure) to simulate real-world operations and define predictive maintenance schedules. Generates digital twin specifications, sensor mapping requirements, and simulation parameters for operational planning. Trigger on queries about digital twins, virtual replicas, predictive maintenance planning, simulation models, sensor mapping, and operational simulation. Do NOT trigger on general IoT device management, dashboard design, data visualization, supply chain analytics without simulation context, or hardware procurement.
development
game-day-simulator
Disaster recovery and chaos engineering tabletop exercise orchestrator. Designs and facilitates structured game day scenarios that test incident response procedures, failover mechanisms, backup restoration, and team communication under simulated outage conditions. Produces scenario scripts with inject timelines, expected vs. actual response matrices, and post-exercise scoring rubrics. Use this skill whenever the user mentions game day, tabletop exercise, disaster recovery drill, chaos engineering scenario, DR test, failover test, incident simulation, business continuity exercise, fire drill for infrastructure, or wants to practice incident response — even if they just say "we need to test our DR plan" or "let's simulate an outage". Do NOT trigger when the user is debugging a real production incident (use incident-response workflow), asking about actual Kubernetes pod failures (use k8s-debugger), performing real infrastructure cost analysis (use cloud-finops-optimizer), or writing runbooks for existing procedures (that's documentation).
development
health-tracker
Parse natural-language health check-ins and log them to ~/health_data.json for dashboard sync. Activate when the user mentions food, water, walking, exercise, yoga, breathing, sleep, or screens.
testing
human-ai-trust-architect
Designs transparency, explainability, and auditability frameworks to ensure humans can meaningfully oversee and audit autonomous AI decisions. Produces trust architecture documents including explanation templates, logging requirements, override mechanisms, and confidence-calibration standards. Trigger on queries about AI trust, explainability frameworks, AI transparency, human oversight, AI auditability, explanation design, and trust architecture. Do NOT trigger on general AI/ML model building, AI ethics policy writing, UI/UX design without trust context, compliance auditing, or data privacy implementation.
development
iac-drift-remediator
Infrastructure-as-Code state and drift remediation specialist. Diagnoses Terraform state file corruption, resolves state drift between declared IaC and actual cloud resources, generates targeted remediation plans using terraform import/state rm/state mv commands, and produces safe rollback procedures. Supports Terraform, OpenTofu, Pulumi, and CloudFormation. Use this skill whenever the user mentions state drift, terraform state, state file corruption, resource import, state lock, terraform plan showing unexpected changes, "wants to destroy and recreate", out-of-band changes, manual cloud console changes breaking IaC, or state migration — even if they don't explicitly say "drift". Do NOT trigger when the user is asking about writing new Terraform modules from scratch (that's IaC authoring), cloud cost optimization (use cloud-finops-optimizer instead), Kubernetes manifest debugging (use k8s-debugger), or CI/CD pipeline configuration.
development
immersive-storyteller
Crafts visual narratives and multimedia content specifically adapted for spatial computing, XR (Extended Reality) experiences, and immersive environments. Designs story arcs, environmental narratives, and interactive content flows optimized for VR/AR/MR platforms where the user is inside the story. Do NOT trigger on traditional video production or screenwriting, 2D web content creation, podcast production, social media content, or general copywriting.
development
k8s-debugger
Kubernetes triage and debugging specialist. Systematically diagnoses pod failures (CrashLoopBackOff, ImagePullBackOff, OOMKilled, Pending), networking issues (DNS resolution, service mesh, ingress), storage problems (PVC binding, mount failures), and RBAC/admission errors. Produces structured triage runbooks with kubectl commands and remediation steps. Use this skill whenever the user mentions CrashLoopBackOff, OOMKilled, ImagePullBackOff, pod pending, kubectl describe, pod logs, Kubernetes networking, service mesh debugging, ingress not working, PVC pending, node not ready, evicted pods, failed deployments, HPA not scaling, or any Kubernetes cluster troubleshooting — even if they just paste a kubectl error output. Do NOT trigger when the user is asking about writing new Kubernetes manifests or Helm charts from scratch (that's authoring, not debugging), cloud cost optimization (use cloud-finops-optimizer), Terraform state issues (use iac-drift-remediator), or CI/CD pipeline failures unrelated to Kubernetes runtime.
development
mutation-test-generator
Deliberately injects bugs (mutations) into code to evaluate test suite strength and generates highly robust unit tests that catch these mutations. Focuses on mutation testing methodology to reveal gaps in test coverage. Identifies mutation operators such as boundary changes, operator swaps, and return value flips to ensure comprehensive test quality. Do NOT trigger: general unit test writing without mutation context, code coverage reporting only, integration or E2E testing, performance or load testing, debugging failing tests
development
performance-bottleneck-debugger
Uses causal reasoning and distributed tracing analysis to locate the root microservice causing unpredictable latency spikes in complex service meshes. Analyzes trace spans, dependency graphs, and timing correlations to pinpoint bottlenecks that traditional APM tools miss. Trigger when: latency spikes occur, performance bottlenecks need identification, slow microservice investigation, trace analysis required, distributed tracing debug needed, p99 latency investigation, service mesh latency issues, or root cause latency analysis. Do NOT trigger for: general application debugging without latency context, Kubernetes pod lifecycle issues, network connectivity troubleshooting, load testing or capacity planning, or cost optimization.
tools
semantic-communications-optimizer
Crafts messaging and content structured specifically for rapid machine-to-machine parsing and AI intent transfer. Optimizes communications for machine readability while maintaining human clarity — ensuring content is parseable by APIs, chatbots, AI agents, and automated workflows. Do NOT trigger on general copywriting or content creation, traditional SEO, social media content, email marketing, or human-only communications without machine context.
development
health-tracker
Track daily health habits (nutrition, exercise, lifestyle) through natural-language check-ins. Parses casual messages about food, water, walks, workouts, sleep, and wellness into structured data for a visual dashboard. Use this skill whenever the user mentions anything about eating, drinking water, walking, exercising, yoga, breathing, sleep, screens, or health habits — even if they don't say 'check-in' or 'log'. Also triggers on first-time setup when someone wants to start tracking their health.
testing
test-skill
One-line description of what this skill does
testing
whimsy-injector
Audits user journeys and product interfaces to identify opportunities for injecting personality, delight, micro-interactions, Easter eggs, and playful moments that reduce task anxiety and increase emotional engagement. Applies the science of surprise and delight without compromising usability or accessibility. Trigger on adding personality, micro-interactions, Easter eggs, reducing task anxiety, and emotional design. Do NOT trigger on general UI/UX design reviews, accessibility auditing, brand voice writing, gamification strategy, or animation/motion design without delight context.
testing
semantic-schema-mapper
Uses semantic analysis and local embeddings to map ambiguous legacy database columns to modern canonical schemas. Resolves naming inconsistencies (e.g., cust_nm → customer.full_name), infers data types from sample values, and generates migration DDL and ETL transformation logic. Trigger when: schema mapping, column mapping, legacy database migration, schema translation, canonical schema, field mapping, database schema alignment, map legacy columns. Do NOT trigger: general SQL query writing, database performance tuning, data visualization, ETL pipeline orchestration without schema context, or data quality checks without mapping needs.
development
chaos-engineering-tuner
Proactively simulates traffic spikes, regional failures, dependency outages, and resource exhaustion to stress-test auto-scaling mechanisms and verify system resilience. Designs chaos experiments with blast radius controls, generates experiment configurations, and produces resilience scorecards. Trigger when: chaos engineering experiments needed, chaos experiment design required, stress testing planned, fault injection scenarios, resilience testing needed, blast radius control, chaos monkey simulation, gameday experiments, LitmusChaos or Chaos Mesh experiments, or auto-scaling verification. Do NOT trigger for: tabletop disaster recovery exercises, load testing for capacity planning only, general performance tuning, infrastructure provisioning, or security penetration testing.
testing
xr-spatial-interface-architect
Designs 3D user interfaces, interaction models, and spatial UX patterns specifically for AR/VR/MR environments. Produces interaction design specs including gaze tracking, hand gesture vocabulary, spatial audio cues, comfort zone mapping, and locomotion patterns that prevent VR sickness. Trigger on XR interface design, spatial UI architecture, VR/AR interaction modeling, and mixed reality UX challenges. Do NOT trigger on traditional 2D web or mobile UI design, general UX research without XR context, immersive storytelling/narrative, game design, or 3D modeling/rendering.
development
multicloud-governance-enforcer
Translates complex regulatory requirements (SOC2, HIPAA, GDPR, PCI-DSS) into compliance-as-code policies to enforce strict access controls, data residency, and encryption standards across AWS, GCP, and Azure environments. Generates OPA/Rego policies, Sentinel rules, and audit reports. Trigger when: compliance as code needed, governance policy enforcement, multicloud compliance required, OPA/Rego policy generation, regulatory compliance implementation, data residency enforcement, access control policies, compliance enforcement, or policy-as-code setup. Do NOT trigger for: general cloud cost optimization, infrastructure provisioning, Terraform drift detection, application security scanning, or identity management setup.
development
accessibility-wcag-auditor
Strictly audits frontend code, UI components, and design mockups against WCAG 2.2 AA standards. Identifies violations in color contrast, keyboard navigation, screen reader compatibility, ARIA attributes, focus management, and touch target sizing. Generates prioritized remediation reports with code fix suggestions. Trigger on queries about WCAG audits, accessibility audits, a11y checks, color contrast, screen reader compatibility, keyboard navigation, ARIA attributes, and accessibility remediation. Do NOT trigger on general UI/UX design feedback, visual design critique, performance optimization, SEO auditing, or cross-browser compatibility testing.
development
inclusive-visuals-specialist
Generates precise image generation prompts and audits AI-generated visual media to guarantee bias-free, diverse, and authentic representation across skin tones, body types, ages, abilities, cultures, and gender expressions. Ensures visual content meets inclusive design standards and avoids harmful stereotypes. Trigger on inclusive visuals, diverse representation, bias-free images, and representation auditing. Do NOT trigger on general graphic design without inclusion focus, cross-cultural UX auditing, accessibility/WCAG auditing, brand identity design, or photography art direction.
testing
Skill Name
One-line description of what this skill does
tools
cognitive-load-balancer
Analyzes team workflows, task dependencies, and context-switching patterns to dynamically reorganize work assignments that reduce mental fatigue and cognitive overhead. Models task complexity, attention cost of switches, and focus-time requirements to optimize human productivity. Trigger on queries about cognitive load, context switching, mental fatigue, workflow optimization, task reorganization, focus time, and attention management. Do NOT trigger on general project management, sprint planning, Jira/Linear ticket triage, team capacity planning without cognitive context, performance reviews, or process documentation.
testing