387 skills · sorted by relevance
Plan and review machine learning systems for production use, covering data requirements, repeatable training, quality checks, deployment, monitoring, and rollback. Produce concrete plans and review findings for moving model experiments into operational systems.
Find guidance on Codex settings, pricing, and OpenAI products and APIs using current official documentation. Get cited answers for setup, troubleshooting, model comparisons, migration planning, and prompt changes, with uncertainty stated when sources are inconclusive.
Generate RuVLLM settings for local language model inference and create MicroLoRA or SONA instances. Adjust their small internal weights using quality scores as feedback, with instances held in server memory for subsequent adaptation calls.
Choose embedding models, which represent text as numerical vectors, for meaning-based search and RAG applications. Compare options for code, prose, and multilingual content, and plan document splitting, preprocessing, caching, and dimension reduction to improve retrieval quality.
Evaluate LLM applications using automated scores, human reviews, and model-based judging. Build repeatable checks to compare models or prompts, assess answer quality, and detect declines in performance before deployment.
Build an end-to-end evaluation pipeline for Python AI applications using pixie-qa and real language-model calls. Define criteria, prepare realistic test data, score application runs, and analyze the evidence to produce a prioritized improvement plan.
Check security audit findings against code evidence and trace how untrusted input moves through a system. Guide a second review, document verification gaps, and assess draft reports against a minimum quality score before delivery.
Prepare training datasets for large language models with GPU-assisted cleaning. Covers duplicate removal, quality filtering, and personal information redaction, with workflows for text, images, video, and audio using NeMo Curator.
Turn a feature into an agent skill or improve an existing one. Define purpose-specific quality criteria, review outputs with models from three providers, then add tests and check that user requests reach the intended skill.
Prepare training datasets for large language models with GPU-assisted cleaning. Covers duplicate removal, quality filtering, and personal information redaction, with workflows for text, images, video, and audio using NeMo Curator.
Configure GSD through one command, covering model choice, research, plan checks, verification, and branching. Additional modes handle detailed workflow settings, external service integrations, and model profiles, which are named groups of model settings.
Audit existing public equity investment models and spreadsheets for formula errors, source support, assumptions, and scenario quality. Receive a readiness verdict, an issue log, and prioritized fixes showing whether the model supports its intended decision.
Assess whether data is trustworthy enough for analysis, dashboards, experiments, or modeling. Check missing values, duplicates, consistency, freshness, and changes over time, then explain the evidence, downstream risks, likely causes, and practical fixes.
Design, debug, and tune OpenRouter API integrations. It checks model capabilities, reasoning settings, output limits, and fallback routes, then validates results against quality, cost, privacy, and performance requirements, including failed attempts.
Use when the user wants to migrate code that calls OpenAI, Gemini/Google AI, or the Anthropic API to Amazon Bedrock — a pure model/SDK rewrite. End-to-end: assesses the codebase, rewrites SDK calls, evaluates output quality against Bedrock, and delivers a ready-to-merge git branch. Also has an access-only mode for checking or enabling Bedrock model access for specific models (e.g. 'request access to Claude on Bedrock', 'which models do I turn on') with no code migration — it checks each model and walks the right access step, then stops. Not for agent runtime selection or agent migration planning (use agent-advisor), nor standalone cost estimates or infra-only migration. The full migration REQUIRES gcp-to-aws OR azure-to-aws installed alongside this skill (gcp-to-aws preferred when both are present); access-only mode needs neither.
The English summary is being prepared. Showing the original description.
Resolve model profile (quality/balanced/budget) at orchestration start and map agents to specific models. Enables cost/quality tradeoffs by selecting appropriate AI models for each agent role.
The English summary is being prepared. Showing the original description.
Collect and assess quality metrics for Simulink models using the metric.Engine API. Use when evaluating model design quality (maintainability, complexity, architecture) or testing completeness (unit testing, SIL, PIL, coverage, pass rates) after edits or test runs. Covers metric discovery, collection, report generation, and diagnostics. Triggers on: "collect metrics", "quality metrics", "metric.Engine", "testing metrics", "maintainability metrics", "SIL metrics", "PIL metrics", "what metrics are available", "assess quality", "coverage metrics", "generate metric report", "diagnose metric errors".
The English summary is being prepared. Showing the original description.
Use when comparing Scenario models on the same brief through MCP: a bake-off with shared prompts and inputs, cost, generation time and quality measured side by side, choosing a default model, testing a new release against the incumbent, or finding which model handles a style, edit, or reference image best. Keywords: model comparison, benchmark, bake-off, A/B test, contact sheet, cost per asset, latency.
The English summary is being prepared. Showing the original description.
Run AI agent and LLM evaluations in CI/CD pipelines — automated quality gates that fail the build when AI output quality drops. Use when someone asks to "test my AI agent", "add evals to CI", "catch prompt regressions", "compare models", "evaluate LLM output quality", "set up AI quality gates", or "benchmark my agent before deploying". Covers eval frameworks (Cobalt, Promptfoo, Braintrust), LLM-as-judge scoring, threshold-based assertions, and GitHub Actions integration.
The English summary is being prepared. Showing the original description.
Test how API-accessible language models respond to attempts to bypass safety filters. The skill covers prompt templates, input transformations, parallel response comparisons, and Hermes configuration, with automated scoring of refusals, answer quality, and speed.
Analyze and modify refusal behavior in open-weight language models without retraining. Use OBLITERATUS to select methods suited to available hardware, compare approaches, and evaluate changes in refusal rates and response quality.
Explore concepts learned by language models using sparse autoencoders, which separate internal activity into a small set of features. Analyze pretrained models, train custom autoencoders, assess reconstruction quality, and experiment with feature based output adjustments.
Build MCP servers, connectors that let AI models use external services through tools. This skill guides API research, tool design, implementation in TypeScript or Python, testing, and evaluation with realistic questions that require multiple tool calls.
Compare Claude, GPT, and Gemini on the same skill or prompt. Get a table of response times, token usage, and costs, with optional AI scoring to help choose a model for the task.