Browse the Elephant Agent skill shelf.

Keeps explicit shell work legible, bounded, and attached to the current workspace thread.

Keeps direct URL reading and text extraction available when a specific page matters more than search results.

Keeps lightweight public-web discovery available for prompts, cron jobs, and follow-up research.

Keeps local retrieval, file search, and nearby context gathering available in-session.

Break ambiguous work into explicit, verifiable steps before implementation when the task is broad enough to benefit from a plan.

Package a code change for review with clear scope, risk notes, and validation evidence so reviewers can move quickly.

Guide for creating effective Elephant Agent skills. Use this skill when users want to create a new skill or update an existing skill that extends Elephant Agent with specialized knowledge, workflows, or tool integrations.

Split work into parallelizable, bounded tracks only when the tasks have disjoint ownership and the coordination cost is justified.

Investigate bugs by reproducing, isolating, and validating a root-cause hypothesis before applying fixes.

Drive behavior changes through a failing test, the smallest working implementation, and then cleanup once the behavior is locked.

Write execution plans that preserve intent, scope, dependencies, and validation without turning into vague prose.

Verify GitHub CLI or connector authentication, identify the active identity, and debug missing permissions before other GitHub operations.

Review pull requests with a bug-first mindset, grounding findings in concrete diff context and repository behavior.

Inspect GitHub repositories, files, and diffs methodically before proposing changes, especially when the task starts from a GitHub URL or PR.

Triage GitHub issues, extract the real ask, and connect each issue to the relevant files, tests, or release surface before acting.

Move cleanly through branch, diff, review, validation, and PR update steps without losing scope or repository-native workflow.

Inspect repository metadata, branches, permissions, and operational state before changing settings or branch structure.

Guides operator-owned delegation to Claude Code when repo work benefits from a second coding lane or an interactive code agent.

Guides operator-owned delegation to Codex when the task fits a bounded coding lane, review pass, or worktree-isolated implementation track.

Complete guide to what Elephant Agent is, how to use it from the CLI, how herd, wake, Personal Model, proactive curiosity, skills, and tools fit together, and how to explain Elephant Agent clearly to users, contributors, or evaluators.

Frames OpenCode as an operator-controlled coding lane for bounded implementation or review work inside a real repository.

Produces clean architecture and system diagrams from actual repo or product truth instead of decorative boxes.

Generate ASCII art using pyfiglet (571 fonts), cowsay, boxes, toilet, image-to-ascii, remote APIs (asciified, ascii.co.uk), and LLM fallback. No API keys required.

"Production pipeline for ASCII art video — any format. Converts video/audio/images/generative input into colored ASCII character video output (MP4, GIF, image sequence). Covers: video-to-ASCII conversion, audio-reactive music visualizers, generative ASCII art animations, hybrid video+audio reactive, text/lyrics overlays, real-time terminal rendering. Use when users request: ASCII video, text art video, terminal-style video, character art animation, retro text visualization, audio visualizer in ASCII, converting video to ASCII art, matrix-style effects, or any animated ASCII output."

Generate professional infographics with 21 layout types and 21 visual styles. Analyzes content, recommends layout×style combinations, and generates publication-ready infographics. Use when user asks to create "infographic", "visual summary", "信息图", "可视化", or "高密度信息大图".

"Generate project ideas through creative constraints. Use when the user says 'I want to build something', 'give me a project idea', 'I'm bored', 'what should I make', 'inspire me', or any variant of 'I have tools but no direction'. Works for code, art, hardware, writing, tools, and anything that can be made."

Shapes sketch-style diagrams and editable visual explanations for flows, product ideas, and architecture discussions.

"Production pipeline for mathematical and technical animations using Manim Community Edition. Creates 3Blue1Brown-style explainer videos, algorithm visualizations, equation derivations, architecture diagrams, and data stories. Use when users request: animated explanations, math animations, concept visualizations, algorithm walkthroughs, technical explainers, 3Blue1Brown style videos, or any programmatic animation with geometric/mathematical content."

"Production pipeline for interactive and generative visual art using p5.js. Creates browser-based sketches, generative art, data visualizations, interactive experiences, 3D scenes, audio-reactive visuals, and motion graphics — exported as HTML, PNG, GIF, MP4, or SVG. Covers: 2D/3D rendering, noise and particle systems, flow fields, shaders (GLSL), pixel manipulation, kinetic typography, WebGL scenes, audio analysis, mouse/keyboard interaction, and headless high-res export. Use when users request: p5.js sketches, creative coding, generative art, interactive visualizations, canvas animations, browser-based visual art, data viz, shader effects, or any p5.js project."

Guides reference-aware website and landing-page work so the result feels intentional, current, and product-grade instead of generic.

Help with lyric writing, composition structure, and AI-assisted music workflows from idea to rough demo.

Handle Google Docs, Sheets, Drive, and related Workspace tasks through the narrowest available API or browser workflow.

Work with Linear issues, projects, and triage flows while preserving status semantics and ownership clarity.

Extract, inspect, and summarize PDFs quickly with lightweight tooling before escalating to heavier OCR or layout workflows.

Create, search, and update Notion pages or databases through the API or a narrow browser fallback when a Workspace lives in Notion.

Recover text from scanned or image-heavy documents before attempting structured analysis or downstream writing tasks.

Build or revise presentation structure, slide copy, and export-ready material for PowerPoint-style decks.

Give Elephant Agent phone capabilities without core tool changes. Provision and persist a Twilio number, send and receive SMS/MMS, make direct calls, and place AI-driven outbound calls through Bland.ai or Vapi.

Search, filter, and summarize arXiv papers with explicit titles, authors, dates, and paper links before drawing conclusions.

Monitor blog or website updates, compare publish dates, and summarize deltas rather than repeating unchanged content.

Build concise, cross-linked wiki-style summaries for models, papers, techniques, or labs when the user wants durable knowledge capture.

Inspect prediction-market questions, prices, and resolution conditions carefully before summarizing or comparing market signals.

Draft research-style outlines, section structure, and evidence-backed prose for papers, whitepapers, or technical notes.

Guides notebook-first analysis with reproducible kernels, inspectable data loading, and explicit promotion paths back into durable code.

PyTorch library for audio generation including text-to-music (MusicGen) and text-to-sound (AudioGen). Use when you need to generate music from text descriptions, create sound effects, or perform melody-conditioned music generation.

Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support

OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross-modal retrieval. Trained on 400M image-text pairs. Use for image search, content moderation, or vision-language tasks without fine-tuning. Best for general-purpose image understanding.

Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and agents with DSPy - Stanford NLP's framework for systematic LM programming

Frames model, prompt, and system evaluation as a reproducible experiment with baselines, datasets, and explicit metrics.

Guides model, dataset, and space workflows around Hugging Face Hub with explicit auth, artifact, and cache assumptions.

Guides model-serving and runtime-inference decisions across local, remote, and packaged deployment paths.

Run LLM inference with llama.cpp on CPU, Apple Silicon, AMD/Intel GPUs, or NVIDIA — plus GGUF model conversion and quantization (2–8 bit with K-quants and imatrix). Covers CLI, Python bindings, OpenAI-compatible server, and Ollama/LM Studio integration. Use for edge deployment, M1/M2/M3/M4 Macs, CUDA-less environments, or flexible local quantization.

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML models as APIs, or running batch jobs with automatic scaling.

Remove refusal behaviors from open-weight LLMs using OBLITERATUS — mechanistic interpretability techniques (diff-in-means, SVD, whitened SVD, LEACE, SAE decomposition, etc.) to excise guardrails while preserving reasoning. 9 CLI methods, 28 analysis modules, 116 model presets across 5 compute tiers, tournament evaluation, and telemetry-driven recommendations. Use when a user wants to uncensor, abliterate, or remove refusal from an LLM.

Guarantee valid JSON/XML/code structure during generation, use Pydantic models for type-safe outputs, support local models (Transformers, vLLM), and maximize inference speed with Outlines - dottxt.ai's structured generation library

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

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

Foundation model for image segmentation with zero-shot transfer. Use when you need to segment any object in images using points, boxes, or masks as prompts, or automatically generate all object masks in an image.

State-of-the-art text-to-image generation with Stable Diffusion models via HuggingFace Diffusers. Use when generating images from text prompts, performing image-to-image translation, inpainting, or building custom diffusion pipelines.

Guides fine-tuning and post-training work with explicit data, objective, hardware, and rollback assumptions.

Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.

Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization

Guides retrieval-store design, indexing, and query behavior for embedding-backed systems without confusing storage with application truth.

Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GPU memory. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism.

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

OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.

Guides webhook-driven integrations, event receivers, and trigger flows without inventing runtime support the current repo does not actually have.

"Run 150+ AI apps via inference.sh CLI (infsh) — image generation, video creation, LLMs, search, 3D, social automation. Uses the terminal tool. Triggers: inference.sh, infsh, ai apps, flux, veo, image generation, video generation, seedream, seedance, tavily"

Manage Docker containers, images, volumes, networks, and Compose stacks — lifecycle ops, debugging, cleanup, and Dockerfile optimization.

Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in, and reading/injecting secrets for commands.

Investigate open-source supply-chain incidents with evidence collection, recovery, IOC tracking, and structured forensic reporting.

OSINT username search across 400+ social networks. Hunt down social media accounts by username.

Open Notes.app and create or update iCloud notes on macOS using memo when possible and AppleScript when direct app automation is more reliable.

Open Reminders.app and create macOS reminders with AppleScript when the user wants follow-ups in Apple's reminders system.

Open Find My on macOS, inspect devices or people when the user explicitly asks, and keep side effects gated behind confirmation.

Read or send Messages.app conversations on macOS through CLI or AppleScript with explicit confirmation before outbound sends.

Work with vault-backed markdown notes in Obsidian, preserving links, filenames, and existing note structure.

Discover, inspect, authenticate, and call MCP servers directly from the terminal through mcporter when a generic connector is not enough.

Work through MCP-native resources and tools first when a connector already exposes the needed context or action surface.

Work with IMAP email from the terminal using Himalaya for inbox reads, draft review, and carefully confirmed outbound sends.

Give the agent its own dedicated email inbox via AgentMail. Send, receive, and manage email autonomously using agent-owned email addresses (e.g. elephant-agent@agentmail.to).

Search and download reaction GIFs through the Tenor API using curl and jq when the user wants lightweight media retrieval.

Set up and run HeartMuLa, the open-source music generation model family (Suno-like). Generates full songs from lyrics + tags with multilingual support.

Generate spectrograms and audio feature visualizations (mel, chroma, MFCC, tempogram, etc.) from audio files via CLI. Useful for audio analysis, music production debugging, and visual documentation.

Fetch YouTube transcripts, summarize videos, and transform the transcript into chapters, notes, threads, or article-ready structure.

Set up a modded Minecraft server from a CurseForge/Modrinth server pack zip. Covers NeoForge/Forge install, Java version, JVM tuning, firewall, LAN config, backups, and launch scripts.

Play Pokemon games autonomously via headless emulation. Starts a game server, reads structured game state from RAM, makes strategic decisions, and sends button inputs — all from the terminal.

Interact with X/Twitter through an official-API CLI flow for search, timelines, mentions, and carefully confirmed posting actions.

Control Philips Hue lights, rooms, and scenes from the terminal through the OpenHue CLI with explicit targeting and safe defaults.

Find nearby places or services with location-aware search while keeping assumptions about distance, hours, and category explicit.