AI & Machine Learning skills

>-

Thread-safe data persistence in Swift using actors — in-memory cache with file-backed storage, eliminating data races by design.

Framework formal de evaluación para sesiones de Claude Code que implementa principios de desarrollo orientado a evals (EDD)

Production machine-learning engineering workflow for data contracts, reproducible training, model evaluation, deployment, monitoring, and rollback. Use when building, reviewing, or hardening ML systems beyond one-off notebooks.

SAM: zero-shot image segmentation via points, boxes, masks.

Songwriting craft and Suno AI music prompts.

Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.

Publish static sites to {slug}.here.now and store private files in cloud Drives for agent-to-agent handoff.

Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram algorithms. Train custom vocabularies, track alignments, handle padding/truncation. Integrates seamlessly with transformers. Use when you need high-performance tokenization or custom tokenizer training.

Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and build multi-step workflows with Guidance - Microsoft Research's constrained generation framework

>-

Manage Apple Notes via memo CLI: create, search, edit.

Expert guidance for Fully Sharded Data Parallel training with PyTorch FSDP - parameter sharding, mixed precision, CPU offloading, FSDP2

Agent payments via Stripe Link — cards, SPT, approvals.

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

GitHub auth setup: HTTPS tokens, SSH keys, gh CLI login.

Axolotl: YAML LLM fine-tuning (LoRA, DPO, GRPO).

TRL: SFT, DPO, PPO, GRPO, reward modeling for LLM RLHF.

OBLITERATUS: abliterate LLM refusals (diff-in-means).

High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.

Search personal knowledge bases, notes, docs, and meeting transcripts locally using qmd — a hybrid retrieval engine with BM25, vector search, and LLM reranking. Supports CLI and MCP integration.

Install and select animated petdex mascots for Hermes.

>

Execute a natural-language browser intent via page-agent (browser_act) when the target is easier to describe than to select — degrades gracefully when page-agent or an OpenAI-compatible LLM provider isn't configured

How to use the {{NAME}} plugin and the data it produces.

Inspect and optionally adjust the agent's runtime state. Use to check the current model or preset, context window, iteration progress and limits, token usage, workspace and tool configuration, subagent status, and request routing metadata such as channel, chat ID, and sender ID; diagnose unavailable capabilities; change allowed runtime settings; or store temporary session scratchpad values.

Test a learner on a single Claude Code tutorial lesson (01-10) with 10 questions, scoring answers and flagging weak spots. Use before, during, or after a lesson. Don't use for whole-tutorial assessment or explaining a topic instead of testing it.

>-

Manus-style persistent file-based planning for AI coding agents: keeps task_plan.md, findings.md, and progress.md on disk so work survives context loss and /clear. Use when asked to plan out, break down, or organize a multi-step project, research task, or any work requiring 5+ tool calls. Supports automatic session recovery after /clear.

Browser automation CLI for AI agents. Use this when asked to test something in a real browser.

CodexBar live QA/e2e testing: run provider usage matrix checks, validate real app config, use Peekaboo for menu proof, use Browser Use/official docs for API spec or logged-in dashboard checks, and handle 1Password credentials safely.

Use when an MCP-connected agent should use Memori tools for targeted recall, summaries, post-compaction briefs, durable memory augmentation, quota checks, signup, feedback, preferences, prior context, or cross-session continuity.

Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B - collaborative MLOps platform

Use when building CLI tools, implementing argument parsing, or adding interactive prompts. Invoke for parsing flags and subcommands, displaying progress bars and spinners, generating bash/zsh/fish completion scripts, CLI design, shell completions, and cross-platform terminal applications using commander, click, typer, or cobra.

Hybrid retrieval primitive for the Compound Vault. Replaces the v1.6 static hot→index→drill read order with contextual-prefix + BM25 + cosine-rerank, modeled on Anthropic's Sept 2024 Contextual Retrieval research (35-49-67% retrieval-failure reduction). Opt-in via `bash bin/setup-retrieve.sh`; feature-detected by wiki-query and autoresearch. Triggers on: retrieve, hybrid retrieval, BM25, rerank, contextual retrieval, search the chunks, chunk search, vault search, semantic search, what chunks match, find relevant passages.

>-

CCPM - spec-driven project management: PRD → Epic → GitHub Issues → parallel agents → shipped code. Use this skill for anything in the software delivery lifecycle: writing a PRD ('write a PRD for X', 'let's plan X', 'scope this out'), parsing a PRD into an epic, decomposing an epic into tasks, syncing to GitHub ('sync the X epic', 'push tasks to github'), starting work on an issue ('start working on issue N', 'let's work on issue N'), analyzing parallel work streams, running standups ('standup', 'run the standup'), checking status ('what's next', 'what's blocked', 'what are we working on'), closing issues, or merging an epic. Use ccpm any time the user is talking about shipping a feature, managing work, or tracking progress — even if they don't say 'ccpm' or 'PRD'. Do NOT use for: debugging code, writing tests, reviewing PRs, or raw GitHub issue/PR operations with no delivery context.

Catalog of the AO (Agent Orchestrator) `ao` CLI: spawning workers, managing sessions and projects, sending messages, previewing pages, and daemon control. Use when using the ao CLI, spawning workers, or managing AO sessions in an AO workspace.

Verify an implementer's diff with an INDEPENDENT, different-vendor sub-agent (diff plus contract only); turn blocking issues into fix-tasks and loop until clean.