Claude Skills by nodetool-ai

Ask which skill or flow fits your situation. A router over the skills in this repo.

Use when exploring or modifying a codebase and you need a cheap structural map of files, directories, imports, exports, or direct members before reading full source.

Guide for writing ast-grep rules to perform structural code search and analysis. Use when users need to search codebases using Abstract Syntax Tree (AST) patterns, find specific code structures, or perform complex code queries that go beyond simple text search. This skill should be used when users ask to search for code patterns, find specific language constructs, or locate code with particular structural characteristics.

Review a diff, branch, or PR along three axes — Correctness (does it work, and does it step on a NodeTool landmine like cross-package imports, MsgPack framing, Zustand subscriptions, ui_primitives, packaged-Electron paths, IPC security?), Standards (does it follow the repo's documented rules, plus a Fowler smell baseline?), and Spec (does it implement what the originating issue asked for?). Runs the axes as parallel sub-agents and reports them side by side. Use when the user wants to review c...

Companion CLIs for Runpod workflows — HuggingFace, GitHub, Docker, and AWS.

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

runpod-flash SDK and CLI for deploying AI workloads on Runpod serverless GPUs/CPUs.

A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.

Implement a piece of work based on a spec or set of tickets.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

Produce a whole campaign in NodeTool rather than one video — entity sheets for a product and model, a set of still campaign frames, several short social cuts and a long film, all on one locked grade. Use when the user asks for a launch kit, a campaign, an asset pack, a content set, "a bunch of assets", or names several deliverables in different aspect ratios from one product. Not for a single video (use product-commercial or ugc-video).

Use NodeTool REST API, WebSocket protocol, Chat API (OpenAI-compatible), workflow execution endpoints, and streaming responses. Use when user asks about API endpoints, WebSocket protocol, how to call the API, build a client, integrate with NodeTool, or stream workflow results.

Create browser automation agents that navigate websites, extract data, fill forms, and perform multi-step web tasks using natural language instructions. Use when user asks to automate browsing, scrape websites with AI, build a web agent, or perform complex browser interactions.

Use NodeTool chat CLI commands, interactive terminal, agent mode, workspace management, and Global Chat features. Use when user asks about chat commands, interactive terminal, chat features, agent mode in chat, or the Global Chat interface.

Create custom NodeTool nodes, implement BaseNode subclasses, use @prop decorators, build node packages with process/genProcess methods, register nodes, handle media refs and secrets. Use when user asks to create a node, add a node type, build a custom node, implement a processor, or extend NodeTool with new functionality.

Deploy NodeTool servers using Docker, SSH, RunPod, GCP, or Supabase. Use when user asks to deploy, host, set up a server, configure Docker, use RunPod/GCP/Supabase, manage deployment.yaml, or configure environment variables for production.

Configure AI model providers (OpenAI, Anthropic, Gemini, Ollama, HuggingFace, FAL, Replicate), set up API keys, choose models by task, run local inference with llama.cpp/MLX. Use when user asks about models, providers, API keys, which model to use, or configure any AI provider.

Set up RAG pipelines, vector indexing, document ingestion, vector search, and knowledge base creation in NodeTool. Use when user asks about RAG, document indexing, vector search, chat with documents, knowledge base, embeddings, or collection management.

Debug NodeTool workflow failures, node errors, performance issues, stuck executions, type mismatches, and deployment problems. Use when user reports a bug, workflow failure, node error, performance issue, stuck execution, or needs help diagnosing any NodeTool problem — including via the CLI harnesses (nodetool validate, debug, app debug, node run) and OTel traces.

Build and edit NodeTool visual workflows using UI tools. Use this skill whenever the user asks to create a workflow, connect nodes, build a pipeline, add nodes to a graph, generate images/video/audio via workflows, or asks about workflow patterns. Also use when the user mentions node types, data flow, or wants to automate any multi-step process in NodeTool.

Direct a polished product film in NodeTool — pack shots, macro hero spots, luxury commercials, brand films on location, with or without talent. Use when the user wants an ad that looks shot on a camera rather than a phone, mentions a pack shot, hero shot, product film, brand spot, cyclorama or studio lighting, or names a product as the subject of a multi-shot video. Not for self-filmed or creator-style pieces (use ugc-video).

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.

Use when you need to resolve an in-progress git merge/rebase conflict.

Runpod CLI to manage your GPU workloads.

Direct video where a written voiceover drives the picture — faceless explainers, narrated b-roll, documentary segments, tutorials, corporate and educational video with no on-camera talent. Use when the user asks for an explainer, a voiceover video, a narrated piece, a faceless video, or says the script comes first. Also use when spoken words must time the cut. Not for talking-head or creator video (use ugc-video).

Configure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.

Direct narrative video in NodeTool — short films, trailers, teasers, single dramatic scenes, music-video segments with a story. Use when the user gives a logline, a genre, characters with dialogue, or asks for a short, a trailer, a scene, or a cinematic piece with a title and a score. Not for ads or product films (use product-commercial) and not for narrated explainers (use script-video).

Shared reference for directing video in NodeTool — the storyboard-to-timeline loop, the exact tool contract, entity casting, how a shot becomes a prompt, permission gating and memory. Load this before acting on any storyboard, script, entity or timeline work, and whenever another NodeTool video skill (ugc-video, product-commercial, video-clone, script-video, short-film, launch-kit, video-workflow) tells you to. Consult it directly when the question is which tool, which argument, or why a rend...

Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

Turn the current conversation into a spec and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.

Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in one file per ticket locally, or native blocking links on a real tracker.

Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.

Direct phone-shot video with a person in it — UGC ads, creator testimonials, review-to-hero spots, comedy sketches, day-in-the-life and ritual pieces, first-person travel vlogs. Use when the user wants a video that looks self-filmed or friend-filmed, mentions UGC, TikTok, Reels, a creator, a testimonial, a talking-head ad, or names a person as the subject of a multi-shot video. Not for polished product films (use product-commercial) or narrated b-roll with no talent (use script-video).

Strip "AI slop" from code and prose in this repo — over-engineered abstractions, defensive checks on trusted paths, narrating comments, dead error handlers, redundant types, useEffect-for-derived-data, raw MUI imports, whole-store Zustand subscriptions, throat-clearing prose. Use when reviewing your own diff before commit, when the user says "unslop this", or when refactoring AI-generated code in `web/`, `electron/`, `mobile/`, or `packages/`. Triggers on TypeScript/React/Zustand/MUI/TanStack...

Rebuild an existing video for a new product or cast in NodeTool — break a reference ad down shot by shot, then reproduce its structure with different content. Use when the user attaches or links a video and asks to copy, clone, remake, reverse-engineer, match, or "do it like this one", or wants to know why a reference ad works. Not for directing an original piece from a brief (use ugc-video or product-commercial).

Build a reusable NodeTool workflow graph that produces video, so the same pipeline can be re-run on new inputs — a UGC factory, a batch shot renderer, a per-SKU ad generator. Use when the user asks for a template, a pipeline, a factory, a batch job, or something they will run again with different inputs, rather than one finished video. Not for directing a single piece, which needs no graph (use ugc-video, product-commercial or short-film).

Plan a huge chunk of work — more than one agent session can hold — as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.

Playwright E2E testing patterns, Page Object Model, configuration, CI/CD integration, artifact management, and flaky test strategies. Use when writing, fixing, or deflaking Playwright tests in this repo — the web E2E suites under web/tests/ (e2e-runner, debug-harness, visual) and the Electron E2E setup.

Fixture pack whose dependency imports a Node builtin

Fixture pack whose one module comes from a clean ESM dependency

Fixture pack whose dependency feature-detects process

Fixture pack whose dependency reads process without a guard

Fixture pack whose dependency throws while initializing

Sign AWS requests (S3 and every other SigV4 service) in a Code node or CodeAct action, with NodeTool's signer running on the host

Parse English dates such as next Tuesday in a Code node or CodeAct action, with chrono-node running on the host

Convert, mix, and measure colors inside a Code node or CodeAct action, with culori running in the guest

Parse and write CSV in a Code node or CodeAct action, with papaparse running on the host behind a generated facade