Skills Hub
Agent skills for individuals and teams
Explore reusable SKILL.md agent instructions in Prompeteer.ai Skills Hub, part of a combined library of 5,000+ professional agent workflows and skills for individuals and teams. Review the source and available security-scan information before copying or downloading a skill. The format follows the open agentskills.io standard. Installation, tool access and supported behavior vary by client.
Create a contextual skill · Create and refine contextual skills · Professional workflows · Workspace tools
Gke AI Troubleshooting Tpu Performance Degradation
The skill diagnoses throughput drops and step-time regressions in GKE Cloud TPU training, identifying whether the cause is hardware, network fabric, or workload resource bottlenecks. It employs ML Diagnostics Workload Monitoring and minute-level Cloud Monitoring metrics to pinpoint TPU duty-cycle reductions of 15% or more. This service benefits data scientists and ML engineers who need to resolve performance degradation without pod crashes or full execution stalls.
Gke AI Troubleshooting Tpu Mxla Hang
The skill diagnoses multi‑slice GKE Cloud TPU training hangs by analyzing Megascale XLA hang logs and Cloud Monitoring latency metrics. It distinguishes compiler or launch divergence from host data stalls and TPU, SparseCore, ICI, or network faults, enabling engineers to pinpoint the root cause of frozen training jobs that emit HANG_DETECTED. This service is ideal for teams experiencing abrupt training stalls, but not for gradual throughput drops or pod preemption scenarios.
Sign In With Google Web
The skill implements, configures, and secures Sign In With Google using Google Identity Services across web architectures, enabling developers to embed Google sign‑in buttons, One Tap, and GIS components in React, Next.js, Angular, or plain HTML. It verifies ID tokens on backend runtimes such as Python, Node.js, Go, and Java, enforces domain restrictions, and protects client‑side tokens with WebCrypto nonces or encrypted IndexedDB. The skill also manages cross‑origin iframe integration, CSP, COOP, and Permissions Policy headers, while handling sign‑out and revocation for modern web applications.
Codonfm Score
Codonfm Score validates, prepares, and executes public CodonFM Encodon masked-codon variant scoring, ensuring compatibility of its scoring workflows. It serves researchers and bioinformaticians who explicitly request CodonFM or Encodon analyses, providing precise variant scoring when the context is established. The skill refrains from generic variant-scoring requests, prompting users for the variant and intended analysis first.
Codonfm Finetune
Fine-tune public CodonFM Encodon checkpoints on labeled coding-sequence or coding-variant data using LoRA, head-only, or full fine-tuning. The skill is employed when a user explicitly requests regression or classification fine-tuning of CodonFM or Encodon. It supports only generic public‑v1 Encodon workflows, rejecting Decodon, MissenseDataset, missense_synom_agg, and generation workflows.
Codonfm Setup
The Codonfm Setup skill installs the public CodonFM v1 repository and downloads Encodon checkpoints, enabling developers to build or launch the CodonFM development container, configure local data and checkpoint mounts, and verify GPU access. It also provides easy access to public Encodon weights such as 80M, 600M, 1B, and Cdwt-1B. This skill is ideal for teams building or deploying CodonFM models, but it does not support Decodon, Encodon 5B/10B, missense-aggregation, or codon-optimization configurations.
Codonfm Embed
Validate coding-sequence CSVs, extract public CodonFM/Encodon embeddings, and choose checkpoints for downstream property modeling.
Google Cloud Scc Remediation
The skill automatically remediates Google Cloud Security Command Center findings, correcting IAM permissions, resource misconfigurations, vulnerabilities, and toxic combinations. It is employed when a specific finding or attack path requires mitigation, providing precise configuration fixes and containment guidance. The skill is not intended for general IAM policy queries without an associated Security Command Center finding.
Nvidia Ontology Management
Model and publish semantic definitions in Auto Ontology. Use for terms, relationships, measures, imports, and governed results—not deployment or querying.
Nvidia Ontology Query
Query Auto Ontology and validate generated SQL, rows, and answers. Use for MCP or REST access, readiness, authentication, conversations, and grounded questions.
Cloud Trace Querying
The skill queries Cloud Trace spans, filtering by latency thresholds or error status, and correlates distributed traces with Cloud Logging. It diagnoses latency bottlenecks across Google Cloud services, helping engineers investigate slow requests, analyze trace hierarchies, and resolve latency regressions. The skill is intended exclusively for GCP telemetry and not for non‑GCP data or database query optimization.
Huggingface Llm Trainer
The Huggingface Llm Trainer enables professionals to train or fine‑tune language and vision models on cloud GPU infrastructure using TRL or Unsloth, supporting SFT, DPO, GRPO, reward modeling, and GGUF conversion for local deployment. It guides users through dataset preparation, hardware selection, cost estimation, and monitoring with Trackio, while handling Hub authentication, model selection, and persistence. This skill is ideal for data scientists and ML engineers who need efficient, cloud‑based training workflows without local GPU resources.
Gke AI Troubleshooting Node Unresponsive Timeout
The skill diagnoses and resolves GKE TPU or GPU nodes that become unresponsive, marked as NotReady or NodeStatusUnknown due to kernel panics, hardware lockups, or disabled auto‑repair. It is applied when nodes cease heartbeating past the auto‑repair threshold and pods linger in Terminating, ensuring swift restoration for production workloads. The skill is not intended for healthy nodes, application‑level crashes, or routine GKE upgrades.
Cost Optimization
The Cost Optimization skill analyzes Azure resources to identify idle or orphaned assets, such as unused disks, deleted VMs, or lingering public IPs, and evaluates reservation and savings plan utilization. It recommends right‑sizing and commitment adjustments to reduce cloud spending. The skill benefits cloud finance and operations teams seeking precise cost reductions without involving forecasting or budgeting.
Discover Azure Skills
The skill searches the Azure skills catalog and recommends installable agent skills by matching an Azure task to skill metadata and plugin installation guidance. It is used before starting any task that involves an Azure or Microsoft‑cloud service, product, or data source when no currently loaded skill or tool already covers it. This enables teams to quickly acquire the precise capabilities they need for cloud operations.
Cost Estimation
The Cost Estimation skill forecasts Azure spend for planned resources or workloads, enabling users to project next‑month costs, compare forecasts to planning targets, and evaluate VM costs, storage tiers, and Azure regions. It also allows comparison of retail prices, EA rate cards, and negotiated price sheets. The skill is designed for financial planning and budgeting, not for budget health alerts, bill analysis, cost spikes, rightsizing, or commitment management.
Cost Governance
The Cost Governance skill enables organizations to manage Azure spending by configuring budgets, setting alerts, applying tags, and enforcing policy restrictions. It provides real‑time visibility into budget health, identifies overruns, and ensures compliance with cost center tagging and approved VM SKUs and regions. The skill is ideal for finance and operations teams seeking proactive cost control without handling forecasting or rightsizing tasks.
Cost Analysis
The skill analyzes actual Azure spend, detecting bill changes and pinpointing high‑cost resources such as AKS clusters, Cosmos DB, and Foundry models. It highlights cost spikes, unexpected charges, and idle capacity, enabling IT finance teams to understand bill increases and isolate the most expensive components. The tool does not provide forecasting, pricing estimates, rightsizing, commitments, or budgeting.
Muse_db
Inspect database-backed Muse records for diagnosis and cross-table tracing when purpose-built product tools do not expose the needed state.
Wearable_device_skills
Use when the user asks to discover, inspect, or invoke an agentic capability dynamically published by a paired phone or wearable, including device controls, app actions, camera or media actions, and smart-home actions.
Forget
The Forget skill erases a specified personal fact, preference, relationship detail, topic, or prior event from Muse’s active memory and prevents any existing copies or automations from reintroducing it. It is used for explicit requests such as “forget that” or “don’t remember this about me.” This capability benefits users who need to ensure sensitive or outdated information is permanently removed from the system.
Share_ideas
Publish a reusable native Muse Idea with idea.share when the user explicitly asks to share or publish an Idea or asks for a muse.ai/ideas link. Never offer it unprompted. Not for pages or write-ups, sending a result to someone, social posts, or sharing an artifact.
Outlook_calendar
View, create, update, and delete events in the user's Outlook Calendar.
Calendly
View Calendly events and event types, and manage scheduling data using the Calendly CLI.