AI & ML

outlinedriven-odin-claude-plugin Bundle

Use when debugging CUDA with cuda-gdb or Compute Sanitizer, reading GPU core dumps, using device printf, or triaging error codes 700, 701, 702, and 719. Not for performance: use cuda-profiling.

outlinedriven-odin-claude-plugin Bundle

Use when profiling CUDA with Nsight Systems or Nsight Compute, reading roofline and occupancy metrics, or annotating phases with NVTX. Not for correctness: use cuda-debugging.

pramoddutta Skill

Evaluate RAG pipelines with Ragas, measuring faithfulness, answer relevancy, context precision and recall, building golden datasets, and wiring threshold gates into CI for retrieval regressions.

tonone-ai Skill

Audit ranking quality — metric trends, failure modes, dataset coverage, reranker performance. Use when asked to "audit our ranking quality", "find search relevance failures", or "review reranker performance".

tonone-ai Skill

Design an evaluation framework for a ML model — metrics, splits, and reporting. Use when asked "how should we evaluate this model", "design evaluation metrics", or "plan our train test split".

tonone-ai Skill

Audit existing fine-tuning or prompt engineering work — find quality gaps and optimization opportunities. Use when asked to "audit our fine-tuning work", "find prompt engineering gaps", or "look for optimization opportunities".

tonone-ai Skill

Audit existing vector search or RAG implementation — find quality gaps and performance issues. Use when asked to "audit our RAG system", "why is retrieval bad", or "find vector search quality gaps".

outlinedriven-odin-claude-plugin Bundle

Use when asked to run /model-retuning to retune a skill corpus for a new model, measurement-first. Not for static audits: use deslop.

pramoddutta Skill

Teaches the agent the right way to mock in Jest — jest.fn, mockImplementation, mockResolvedValue, jest.mock factories, spyOn with restore, and isolating modules like axios.

pramoddutta Skill

Streaming API testing skill covering Server-Sent Events testing, chunked transfer encoding, gRPC streaming, real-time data validation, backpressure testing, connection resilience, and AI/LLM streaming response testing.

outlinedriven-odin-claude-plugin Bundle

Use when a user requests an independent Codex or Gemini review of uncommitted code, a branch diff, or a specific commit. Don't use for unconfirmed or unattended review sends, or source mutation.

outlinedriven-odin-claude-plugin Bundle

Use when the user runs /skill-benchmark to score skills, compare models, or gate a skill release. Modes: score (default) and gate. Not for editing skills: use agent-surface-forge.

tonone-ai Skill

Evaluate model performance — check for accuracy drops, data drift, and error patterns. Use when asked about "model accuracy dropped", "evaluate the model", "check for drift", or "model performance".

tonone-ai Skill

Design drift alerts and escalation — thresholds, runbooks, and retrain triggers. Use when asked to "alert on model drift", "when should we retrain", or "write a drift escalation runbook".

tonone-ai Skill

Audit embedding infrastructure — model drift, index freshness, query latency, coverage gaps. Use when asked to "audit our embeddings", "is our index stale", or "find embedding coverage gaps".

tonone-ai Skill

Audit existing eval coverage — gaps, metric validity, benchmark leakage, dataset freshness. Use when asked "what eval coverage do we have", "audit our evals", or "check for benchmark leakage".

tonone-ai Skill

Design and build database schema — tables, columns, types, indexes, constraints, relationships. Given a domain description, output the schema and write the files. Use when asked to "design schema", "database design", "create tables", or "data model".

pramoddutta Skill

Author effective Cursor rules in .cursor/rules/*.mdc - YAML frontmatter (description, globs, alwaysApply), the four rule types, scoped QA rules, subagents, and structure that actually steers the model.

pramoddutta Skill

Gate RAG pipelines in CI with versioned golden eval sets, per-metric thresholds, baseline drift detection, and a build that fails when retrieval or answer quality regresses.

outlinedriven-odin-claude-plugin Bundle

Use when analyzing warp divergence, memory coalescing, shared memory bank conflicts, cache behavior, atomics, or occupancy tradeoffs on NVIDIA and AMD GPUs. Not for tool commands: use cuda-profiling.

outlinedriven-odin-claude-plugin Bundle

Use when asked to improve, optimize, rewrite, tune, or port a prompt, skill, or tool description. Also for building prompt evals or auditing prompt text for dated instructions.

tonone-ai Skill

Design a customer health scoring model — define signals, weights, thresholds, and action triggers. Use when asked to "build health scoring", "how do we predict churn", "what signals indicate a customer is at risk", or "design our health model".

pramoddutta Skill

Test LLM applications with DeepEval, pytest-style unit tests for LLM outputs using G-Eval, answer relevancy, faithfulness, hallucination and custom metrics, with CI quality gates and dataset-driven regression runs.

pramoddutta Skill

Make Claude Code write and maintain end-to-end tests like a senior SDET — Playwright and Cypress flows with stable locators, the Page Object Model, fixtures, reused auth state, network mocking, and flake-free CI. Claude Code E2E testing, done right.

fdu-ins Skill

Use the corp-finance-mcp server tools for specialty finance, regulatory, and compliance calculations. Invoke when performing private credit (unitranche, direct lending, syndication), insurance (loss reserving, premium pricing, Solvency II SCR), FP&A (variance analysis, break-even, working capital, rolling forecast), wealth management (retirement planning, tax-loss harvesting, estate planning), restructuring (recovery analysis, distressed debt), real assets (property valuation, project finance), venture capital (dilution, convertible instruments, fund returns), ESG (scoring, climate/carbon, green bonds, SLL), regulatory capital (Basel III, LCR/NSFR, ALM), compliance (MiFID II best execution, GIPS reporting), credit derivatives (CDS pricing, CVA/DVA), convertible bonds (binomial tree pricing, scenario analysis), lease accounting (ASC 842/IFRS 16, sale-leaseback), pension & LDI (funding analysis, liability-driven investing), sovereign risk (bond analysis, country risk), real options (binomial valuation, decision

outlinedriven-odin-claude-plugin Bundle

Use when asked for a call graph, class hierarchy, dependency map, containment or complexity view, or data-flow view. Not for embedding: use embed-diagram. Not for architecture: use visual-diagram.

outlinedriven-odin-claude-plugin Bundle

Use when the user runs /workspace-unfreeze on a frozen path to make it editable again. Not for automated or unattended runs: requires explicit human invocation.

pramoddutta Skill

Teach agents to use AI browser agents for exploratory and smoke QA with step budgets, evidence-based assertions, guardrails, and Playwright conversion.

pramoddutta Skill

Effective patterns for using GitHub Copilot to generate, refactor, and maintain test code including prompt engineering for test generation, Copilot Chat for debugging, inline suggestions for assertions, and workspace context optimization.

tonone-ai Skill

Audit existing model evaluation code — find metric misuse, missing CIs, and evaluation leakage. Use when asked to "audit our model evaluation", "find metric misuse", or "check for evaluation leakage".

pramoddutta Skill

Teaches the agent when and how to use page.evaluate, evaluateHandle, and exposeFunction in Playwright — passing arguments safely, reading DOM/JS state, and why locators should be preferred for actions.

pramoddutta Skill

Teach AI coding agents to use the Playwright CLI and debug loop efficiently with last-failed runs, locator probing, trace evidence, and safe healing.

pramoddutta Skill

Teaches the agent to build resilient Playwright locators with .filter (hasText/has/hasNot), narrow lists, and reason correctly about visibility, waitFor states, and timeouts.

tonone-ai Skill

Debug AI system behavior using traces — prompt reconstruction, output comparison, failure attribution. Use when asked "why did the model output this", "debug this LLM failure", or "reconstruct the prompt".

tonone-ai Skill

Audit LLM observability coverage — trace gaps, logging completeness, cost attribution accuracy. Use when asked to "audit our LLM observability", "find trace coverage gaps", or "check cost attribution accuracy".

tonone-ai Skill

Systematically optimize prompts for a task — few-shot, chain-of-thought, structured output. Use when asked to "optimize this prompt", "add chain-of-thought", or "improve structured output".