AI & ML

j4flmao Skill

Agentic Workflows & Multi-Agent Orchestration

j4flmao Bundle

Use this skill when asked about data versioning, DVC, LakeFS, data lineage, Git-like for data, data reproducibility, data branching, data diff, data snapshot, or experiment reproducibility. This skill enforces: DVC patterns for ML pipelines and data version control, LakeFS for Git-like semantics on data lakes, branching and merging data, data diff and rollback, experiment reproducibility with metrics tracking, and data lineage from source to model. Do NOT use for: code versioning with Git, database schema migrations, or application artifact versioning.

j4flmao Bundle

Use this skill when working with Prisma ORM — schema modeling, migrations, queries, relations, middleware, and deployment. This skill enforces: data model normalization, relation conventions, Prisma Client best practices, migration safety, and performance optimization. Requires @prisma/client. Do NOT use for: Mongoose, TypeORM, Drizzle ORM, or non-relational databases.

j4flmao Skill

Act and think like a Staff-level AI Solutions Architect, focusing on RAG vs Fine-tuning tradeoffs, cost optimization, and agentic patterns.

j4flmao Bundle

Use this skill when the user says 'biometrics', 'Face ID', 'Touch ID', 'fingerprint', 'fingerprint auth', 'biometric auth', 'face unlock', 'local authentication', 'biometric prompt', 'device credential'. Implement biometric authentication with secure storage on iOS and Android. Do NOT use for: server-side authentication or password-based auth.

j4flmao Bundle

Use this skill when designing experiments, A/B tests, multi-armed bandits, randomized controlled trials, statistical hypothesis tests, sample size calculations, experiment design, power analysis, or causal inference for product changes. This skill enforces: rigorous experiment design with pre-registered hypotheses, proper sample size calculations, statistical significance testing, multiple comparison corrections, guardrail metrics, and result interpretation. Do NOT use for: observational causal inference (see causal-inference skill), general statistical analysis (see statistical-analysis), or ML model evaluation.

j4flmao Bundle

Use this skill when asked about causal inference, causal effect estimation, potential outcomes, Rubin causal model, DAGs, Pearl do-calculus, counterfactual reasoning, structural causal models, difference-in-differences, regression discontinuity, instrumental variables, propensity score matching, synthetic control, uplift modeling, heterogeneous treatment effects, CATE estimation, meta-learners, causal forests, double ML, or causal machine learning. This skill enforces: causal frameworks (potential outcomes, DAGs, do-calculus, counterfactuals, structural causal models), quasi-experimental methods (DiD, RDD, IV, PSM, synthetic control), and causal ML (uplift modeling, CATE, meta-learners, causal forests, double ML). Do NOT use for: A/B testing (use experimentation skill), general statistical analysis (use statistical-analysis skill), or predictive ML modeling.

j4flmao Skill

Rules for breaking down complex projects into granular checklists in task.md, using specific notation, and updating frequently for transparency.

j4flmao Skill

Advanced design of Retrieval-Augmented Generation, focusing on multi-hop retrieval and query rewriting.

j4flmao Bundle

Use this skill when asked about Ethereum internals, EVM deep dive, Ethereum consensus layer, execution clients, staking, EIPs, layer-2 scaling, account abstraction, PBS/MEV-Boost, EVM opcodes, and Ethereum protocol development. Languages: Go, Rust, C#, C++, Solidity. Covers EVM architecture (opcodes, gas metering, memory/storage model, EOF), execution clients (geth, reth, Nethermind, Erigon), consensus layer (Casper FFG, LMD-GHOST, beacon chain, attestation), staking and validators (32 ETH, withdrawal credentials, MEV-Boost, DVT, ePBS), account abstraction (ERC-4337, EntryPoint, paymasters, UserOp mempool), critical EIPs (1559, 4844, 4337, 3529, 2718), and L2 scaling (Optimism, Arbitrum, ZKsync, StarkNet, validium, data availability). Do NOT use for: Bitcoin protocol (use blockchain-bitcoin), non-EVM blockchains (use blockchain-core), or smart contract development (use blockchain-application).

j4flmao Skill

Use this skill when the user says 'CQRS', 'command query segregation', 'separate read write model', 'command model', 'query model', 'read model', 'write model', 'materialized view', 'command handler', 'query handler'. This skill enforces: strict command/query separation, write model optimized for consistency, read model optimized for performance, eventual consistency between models, command validation before execution. Applies to any backend stack. Do NOT use for: simple CRUD applications, event sourcing (use event-sourcing skill), or microservices decomposition (use microservices skill).

j4flmao Skill

First principles of AI societal structures and multi-agent interaction paradigms.

itsual Skill

Apply compliance controls to AI and automated decision systems — transparency, fairness, human oversight, documentation, and regulatory alignment. Use when deploying AI in regulated processes (credit, fraud, customer decisions) or responding to AI governance expectations.

itsual Skill

Deliberately manage what context is loaded, retained, and passed to the model for better reasoning and lower noise. Use when working on large codebases, long sessions, or when the agent seems to be missing important information or getting distracted by irrelevant details.

itsual Skill

Govern model risk — inventory, validation, monitoring, and challenge — for credit, pricing, AML, and other decision models. Use when establishing model risk management (MRM) frameworks or preparing models for regulatory scrutiny.

itsual Skill

Apply Domain-Driven Design tactical and strategic patterns to model complex business domains clearly. Use when building complex business logic, defining bounded contexts, aggregates, or ubiquitous language.

itsual Skill

Protect your domain model from external systems by translating between external models and your own. Use when integrating with legacy systems, third-party APIs, or other bounded contexts that should not leak into your core model.

itsual Skill

Conduct multi-axis code review covering correctness, readability, architecture, security, and performance. Use before merging any change, after implementing a feature, when reviewing code written by yourself, another agent, or a human, or when assessing overall code quality. Trigger on any request involving code review, PR review, quality check, or "review this change".

j4flmao Bundle

Use this skill when performing statistical analysis, hypothesis testing, regression analysis, time series forecasting, Bayesian inference, descriptive statistics, data exploration, or any general statistical modeling. This skill enforces: proper data exploration before modeling, assumption checking, effect size reporting, confidence intervals alongside p-values, robust methods when assumptions violated, and reproducible analysis workflows. Do NOT use for: A/B testing (see experimentation skill), causal inference (see causal-inference skill), or ML model building.

j4flmao Skill

Instructions for creating stunning walkthrough reports using advanced Markdown features.

j4flmao Skill

Local SLM Inference & KV Cache Offloading

j4flmao Skill

AI Agent Swarms: Multi-Agent Debate & Context Routing

j4flmao Skill

Forces the agent to research and write a structured implementation plan before writing any code.

j4flmao Skill

Adversarial AI & Prompt Injection

j4flmao Skill

Advanced RAG Architecture: Algorithmic Foundations and Compilational Paradigms

j4flmao Skill

Academic breakdown of latent reasoning and token probability manipulation.

j4flmao Bundle

Breadth-first search over LLM states and state evaluation reference.

j4flmao Skill

Authoritative guide to writing custom OpenAI Triton kernels for GPU inference optimization, focusing on block-level programming and auto-tuning.

j4flmao Skill

Academic reference for Shor's algorithm period finding and Grover's diffusion operator.

j4flmao Skill

Model Quantization Mechanics: Precision Reduction and Weight Formatting

j4flmao Skill

Master the PagedAttention algorithm and virtual memory management for KV Cache, eliminating fragmentation to maximize LLM throughput.

j4flmao Bundle

Comprehensive skill for designing, implementing, and operating durable execution environments with microVM isolation, snapshotting, forking, and secure agent runtime sandboxes. Covers the sandbox-as-a-tool pattern, Temporal/durable execution frameworks, gVisor/Kata Containers/Firecracker isolation, workspace management, state persistence across sessions, filesystem sandboxing, network isolation, and resource quota enforcement for production agent systems.

j4flmao Bundle

Deep dive into in-context learning mechanics and KV Cache impact.

j4flmao Skill

CUDA Programming & Hardware Acceleration

itsual Skill

Apply Command Query Responsibility Segregation to separate read and write models when their needs diverge significantly. Use when read and write performance, scaling, or model complexity requirements are very different.

itsual Skill

Create, improve, and evaluate new Agent Skills. Use when the user wants to build a new skill from scratch, refine an existing skill, write better descriptions for triggering, add evaluation cases, or package a skill. This is the meta-skill for extending the skills system itself.