Claude Skill Registry

Run GitHub Actions workflows locally with act, including installation, configuration, event payloads, secrets, and debugging.

Apply when tempted to ask 'should I do X?' on reversible work. Proceed, present the result, let the human course-correct after the fact; reserve confirmation for irreversible actions.

Manages Android devices and Wear OS watches via ADB scripts for screenshots, screen recording, tile management, package operations, and device inspection.

N coordinated agents on shared task list with SQLite-based atomic claiming

Provides guidelines for building with Botpress's Agent Development Kit (ADK), covering primitives, project structure, CLI usage, and documentation lookups.

Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a thermo-nuclear code quality review, thermonuclear review, deep code quality audit, or especially harsh maintainability review.

Manages Architecture Decision Records (ADRs) throughout their lifecycle, including reading, writing, and aligning development with documented decisions.

Guide Test-Driven Development using Kent Beck's Red-Green-Refactor cycle. Use when writing tests, implementing features via TDD, or following plan.md test instructions.

Fetches and applies Google API Improvement Proposals (AIP) standards when designing or reviewing APIs, protobuf definitions, or related work.

Remove AI-generated code slop from a branch. Use when cleaning up AI-generated code, removing unnecessary comments, defensive checks, or type casts. Checks diff against main and fixes style inconsistencies.

Decode AKM (Asahi Kasei Microdevices) part numbers, including series, package, interface, and resolution, with guidance for identifying compatible replacements.

Apply to multi-step work (sweeps, migrations, runs of similar edits) and to how you stack commits and PRs. Break work into small units that each end in a verifiable state, check each before the next, and order delivery so the sequence proves itself to a reviewer.

Conduct Gage R&R studies and validate measurement systems per AIAG MSA manual, covering variable and attribute studies, acceptance criteria, and calculation methods.

Orchestrate via @babysitter. Use this skill when asked to babysit a run, orchestrate a process or whenever it is called explicitly. (babysit, babysitter, orchestrate, orchestrate a run, workflow, etc.)

Annotates single-cell RNA-seq data by scoring marker genes, transferring labels with CellTypist, or reasoning over marker lists with an LLM.

Apply when designing commands, lifecycle steps, or processing loops that run amid crashes, restarts, and retries. Converge to the same end state regardless of partial prior runs.

Provides statistical process control guidance including control chart selection, capability indices, and out-of-control rules for quality management.

Apply during planned rewrites and migrations with explicit phase boundaries. Converge on the target architecture; don't preserve smooth intermediate states with throwaway compatibility code.

Coordinates blockchain operations on the Aptos testnet via MCP tools, including balance checks, transfers, swaps, staking, and approvals.

Apply when you catch yourself writing the same instruction a second time, or notice a recurring correction. Encode the rule as a lint, metadata flag, runtime check, or script instead of more text.

Guides building original LangGraph agents for Warden Protocol, from setup to deployment and publishing in Warden Studio.

Apply when designing types, reviewing a function signature, or writing code in any statically-typed language. Make illegal states unrepresentable, brand semantic primitives, parse external data at boundaries, refuse to lie to the compiler, exhaust variants, derive from authoritative schemas.

Short alias that routes invocations to the vc-phase-loop skill for hands-off completion.

Apply when sequencing an addition, refactor, or rewrite. Remove dead code, redundant validators, and stub references first, then build on the simpler base.

Quantize large language models to 4-bit precision using activation-aware weight quantization, reducing memory footprint and speeding up inference with minimal accuracy loss.

Apply when facing a novel UI interaction or architectural decision with no precedent in the codebase. Build 2-3 competing prototypes and compare side by side before committing.

Classifies HTML pages, forms, and fields using machine learning to detect page types, form types, and field types from HTML content or URLs.

Transforms technical documents into rigorous learning journeys with collegiate-level mastery requirements. Uses Bloom's taxonomy progression, 80%+ mastery thresholds, and multi-level verification before advancing. Treats learning as a high school to college graduation progression. Use when user wants deep understanding, not surface familiarity.

Trains language models with Direct Preference Optimization using preference pairs, covering DPOTrainer setup, dataset preparation, and beta tuning for stable preference learning without explicit reward models.

Use when asked to create or refresh a personal mode skill and open a reviewable PR. Mines recent session history for a recurring manual workflow, drafts one self-contained skill file, and opens a PR with evidence. Not for shared repo skills or work that skips human approval.

Generates and analyzes proteins using ESM3 and ESM C language models, covering sequence generation, structure prediction, inverse folding, embeddings, and function conditioning with local or cloud-based Forge API inference.

Generate images, videos, audio, and more using fal.ai AI models, with support for text-to-image, image-to-video, text-to-speech, speech-to-text, image editing, upscaling, model search, workflow creation, and cost estimation.

Query the memory system for relevant learnings from past sessions

Quantize large language models to 8/4/3/2/1-bit precision without calibration data, using multiple optimized backends and integrations with HuggingFace Transformers, vLLM, and PEFT/LoRA.

Apply when reviewing or shaping code that's hard to trace. Count layers between question and answer, and hidden state in the reader's head; collapse one-caller wrappers and shrink mutable scope.

High-performance numerical computing with JAX, covering functional transformations, Flax NNX, and best practices for ML research.

Use when non-trivial code needs a design, architecture needs improving, or one module needs targeted interface narrowing, seams, or testability. Not for diagrams, deploy, or irreversible changes.

Debriefs recent work by probing understanding of the codebase with four questions, tracking concepts in a local ledger to help retain ownership.

Apply when you write, change, or keep a test. Call the code the way its users do and assert the result they observe against a literal expected value. If the test would still pass when every imported function returns undefined, rewrite the assertion or delete the test.

Builds Retrieval-Augmented Generation systems with document chunking, embedding generation, vector storage, and retrieval pipelines, including evaluation and optimization.

Writes technical blog posts about features being built. Triggers when user asks to write about development progress, implementations, or project updates.

Scales Python ML workloads across clusters using Ray's distributed tasks, actors, data, and serving capabilities.

Apply when concurrent actors might write to the same file, branch, key, or state object. Eliminate the sharing first; serialize structurally only when one shared writer is a real invariant.

Annotates single-cell RNA-seq data by scoring marker genes, transferring labels with CellTypist, or reasoning over cluster markers with an LLM.

Resolve git merge conflicts in dependency files, source code, and configuration. Use when merge conflicts are detected.

Fine-tune instruction-following LLMs with Unsloth's optimized SFTTrainer, covering dataset formatting, chat templates, training configuration, and thinking-model patterns.

Apply when introducing a new internal API while old callers still exist. Migrate callers and delete the old API in the same wave instead of preserving compatibility layers.

Search and retrieve book details, author information, and bibliographies via the Open Library API from the command line.

Provides Python code patterns for making HTTP requests with httpx, covering sync and async clients, authentication, streaming, timeouts, retries, and WebSocket support.

Build, deploy, and interact with Juicebox V5 projects, including revnets, hooks, and omnichain deployments, with API references, implementation details, UI generation, and GraphQL queries.

Generates type-safe TypeScript clients from OpenAPI specs using Orval, covering configuration, React Query/SWR hooks, custom mutators, and MSW mocks.

Manage GraphQL schemas and federation with Apollo Rover CLI, covering publishing, fetching, composing, validating, and local supergraph development.

Reference for writing idiomatic, safe Bash and shell scripts with quoting, conditionals, loops, pipes, functions, and error handling.

Runs the CAAM operator skill using Codex and the local shell, following the source skill's domain behavior and output expectations.

Runs the casr operator skill in a Codex environment, reading the source skill for behavior and using the local shell with JSON/NDJSON command surfaces.

Runs the CASS operator skill in a Codex environment, using the local shell and referencing the source skill for domain behavior.

Guides using the Fish interactive shell, covering autosuggestions, universal variables, web configuration, and best practices for interactive use.

Scaffolds a coder-eval task suite in a repository by scanning for testable components, writing one real task, and validating it.

Mines SpecStory coding histories from any agent into a persistent corpus, surfaces reproducible workflows with corroborated evidence, and interactively forges chosen ones into skills installed across agent harnesses.

Persists context across sessions by compiling sources into a local SQLite store and markdown tree, then retrieves relevant memory via embedding search.