Latest Agent Skills

Write ML papers for NeurIPS/ICML/ICLR: design→submit.

Extract text from PDFs/scans (pymupdf, marker-pdf).

Gmail, Calendar, Drive, Docs, Sheets via gws CLI or Python.

Dark-themed SVG architecture/cloud/infra diagrams as HTML.

54 real design systems (Stripe, Linear, Vercel) as HTML/CSS.

Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.

Infographics: 21 layouts x 21 styles (信息图, 可视化).

Create, read, edit .pptx decks, slides, notes, templates.

Design one-off HTML artifacts (landing, deck, prototype).

Edit PDF text/typos/titles via nano-pdf CLI (NL prompts).

Airtable REST API via curl. Records CRUD, filters, upserts.

Monitor blogs and RSS/Atom feeds via blogwatcher-cli tool.

Manim CE animations: 3Blue1Brown math/algo videos.

ASCII video: convert video/audio to colored ASCII MP4/GIF.

Query Polymarket: markets, prices, orderbooks, history.

Install and select animated petdex mascots for Hermes.

Notion API + ntn CLI: pages, databases, markdown, Workers.

Hand-drawn Excalidraw JSON diagrams (arch, flow, seq).

Humanize text: strip AI-isms and add real voice.

Author/validate/export Google's DESIGN.md token spec files.

ASCII art: pyfiglet, cowsay, boxes, image-to-ascii.

Karpathy's LLM Wiki: build/query interlinked markdown KB.

Geocode, POIs, routes, timezones via OpenStreetMap/OSRM.

Use when building creative browser demos with @chenglou/pretext — DOM-free text layout for ASCII art, typographic flow around obstacles, text-as-geometry games, kinetic typography, and text-powered generative art. Produces single-file HTML demos by default.

Generate images, video, and audio with ComfyUI — install, launch, manage nodes/models, run workflows with parameter injection. Uses the official comfy-cli for lifecycle and direct REST/WebSocket API for execution.

Throwaway HTML mockups: 2-3 design variants to compare.

Search arXiv papers by keyword, author, category, or ID.

p5.js sketches: gen art, shaders, interactive, 3D.

Run an AI Startup division meeting with the seven AI Startup agents (no AI Lab agents). Use when the user requests to "start a startup meeting", "convene the startup team", or "run a sprint review for the product".

Architecture specialist for model and training systems. Use when the user asks to talk to Chamber, requests the ML architect, or needs model architecture decisions.

Data pipeline specialist for ML experiments. Use when the user asks to talk to Gekko, requests the data engineer, or needs DataLoader optimization.

Standard reviewer for correctness and reproducibility. Use when the user asks to talk to Omen, requests a code review, or needs reproducibility verification.

Adversarial reviewer that stress-tests claims and conclusions. Use when the user asks to talk to KAY/O, requests an adversarial review, or needs claims validated before publication.

Fast-executing ML engineer for experiments. Use when the user asks to talk to Jett, requests the ML engineer, or needs experiment implementation.

MLOps lead for deployment, monitoring, and scaling. Use when the user asks to talk to McGonagall, requests deployment help, or needs CI/CD for AI systems.

AI product architect guiding PRD-equivalent planning, architecture, and delivery planning decisions. Use when the user asks to talk to Dumbledore, requests the AI architect, or needs structured planning for an AI product.

Run multi-agent research discourse session. Use when the user requests to "start a research party" or "run a journal club".

Implementation specialist for LLM applications and AI systems. Use when the user asks to talk to Hermione, requests the AI engineer, or needs to build an LLM app or agent system.

Run an AI Lab division meeting with research and build agents (no AI Startup agents). Use when the user requests to "start a lab meeting", "convene the lab", or "run a sprint retrospective for the lab".

Data pipeline specialist for AI system integration. Use when the user asks to talk to Hagrid, requests data integration, or needs embedding pipeline or vector DB setup.

AI security and safety specialist for guardrails and adversarial resilience. Use when the user asks to talk to Snape, requests a safety audit, or needs guardrails design.

AI QA and evaluation specialist for quality gates. Use when the user asks to talk to Moody, requests AI evaluation, or needs regression tests for AI systems.

Systems ML and hardware-aware optimization specialist. Use when the user asks to talk to Killjoy, requests the systems researcher, or needs compute-aware architecture tradeoffs, distributed training, and inference optimization.

Prompt and interaction designer for reliable AI behavior. Use when the user asks to talk to Luna, requests prompt engineering, or needs agent behavior specification.

Dataset analysis and data quality specialist. Use when the user asks to talk to Cypher, requests the data detective, or needs dataset assessment, bias analysis, and benchmark evaluation.

Experimental methodology and statistical rigor specialist. Use when the user asks to talk to Breach, requests the methodologist, or needs experiment design review, ablation planning, and reproducibility protocols.

Adversarial robustness and ML safety specialist. Use when the user asks to talk to Viper, requests the adversary, or needs failure mode analysis, attack surface review, and robustness evaluation.

Interdisciplinary synthesis and cross-domain transfer specialist. Use when the user asks to talk to Astra, requests the synthesizer, or needs cross-field insight connections and multi-modal research directions.

Literature survey and SOTA tracking specialist. Use when the user asks to talk to Sova, requests the research tracker, or needs literature review, citation mapping, and paper analysis.

Provides PhD-level mathematical analysis for machine learning, including optimization theory, statistical learning theory, and convergence proofs.

Frontier research tracking and trend scouting specialist. Use when the user asks to talk to Fade, requests the scout, or needs arxiv monitoring, paradigm shift detection, and emerging method analysis.

Run a cross-division all-hands meeting that includes both AI Lab and AI Startup agents. Use when the user requests to "start an all-hands", "convene both divisions", or "run a cross-division meeting".

Shared orchestrator for AI Lab and AI Startup divisions. Use when the user asks to talk to Nosh, requests the ML orchestrator, or says "what should we do next".

Sets up the BMad ML module in a project. Use when the user requests to 'install ML module', 'configure BMad ML', or 'setup BMad ML'.

High-performance toolkit for genomic interval analysis in Rust with Python bindings. Use when working with genomic regions, BED files, coverage tracks, overlap detection, tokenization for ML models, or fragment analysis in computational genomics and machine learning applications.

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation

Bayesian modeling with PyMC. Build hierarchical models, MCMC (NUTS), variational inference, LOO/WAIC comparison, posterior checks, for probabilistic programming and inference.

Fast CLI/Python queries to 20+ bioinformatics databases. Use for quick lookups: gene info, BLAST searches, AlphaFold structures, enrichment analysis. Best for interactive exploration, simple queries. For batch processing or advanced BLAST use biopython; for multi-database Python workflows use bioservices.

Distributed computing for larger-than-RAM pandas/NumPy workflows. Use when you need to scale existing pandas/NumPy code beyond memory or across clusters. Best for parallel file processing, distributed ML, integration with existing pandas code. For out-of-core analytics on single machine use vaex; for in-memory speed use polars.