Trouvez la capacité idéale pour votre agent.
Create distinctive, production-grade frontend interfaces with high design quality. Generates creative, polished code that avoids generic AI aesthetics. Use when the user asks to build web components, pages, artifacts, posters, or applications, or when any design skill requires project context. Call with 'craft' for shape-then-build, 'teach' for design context setup, or 'extract' to pull reusable components and tokens into the design system.
Guide secure smart contract development with best practices for Solidity, Rust (Solana/CosmWasm), and Move. Use when writing, reviewing, or hardening smart contracts against common vulnerability classes like reentrancy, integer overflow, access control issues, and flash loan attacks.
The agent creates and manages fuzzing dictionaries containing domain-specific tokens, magic bytes, protocol keywords, and format-specific strings to guide mutation-based fuzzers past early validation checks. It generates dictionary entries from header files, binary strings, man pages, and LLM prompts, and passes them to libFuzzer via -dict=, AFL++ via -x, or cargo-fuzz. The agent applies this technique when fuzzing parsers (JSON, XML, config files), protocol handlers (HTTP, DNS), file format processors (PNG, PDF, media codecs), or when coverage plateaus indicate the fuzzer cannot discover keyword-guarded code paths without token hints.
Generate or edit images via Gemini 3 Pro Image (Nano Banana Pro). Use when the user asks to create an image, generate a picture, produce AI-generated artwork, edit a photo, compose multiple images, or upscale an image to higher resolution. Supports text-to-image generation, single-image editing, and multi-image composition using the Gemini API.
Extract frames or short clips from videos using ffmpeg. Use when the user asks to grab a frame, capture a screenshot from a video, extract a thumbnail, pull a still image from footage, or snapshot a specific timestamp in a video file.
Gemini CLI for one-shot Q&A, summaries, and generation. Use when the user asks to query Gemini, generate text with Google AI, summarize content using Gemini, run a one-shot prompt, get a Gemini response, or produce JSON output via the Gemini CLI. Supports model selection, output formatting, and extension management.
Batch-generate images via the OpenAI Images API using DALL-E 2, DALL-E 3, or GPT image models. Produces random-but-structured prompts, renders them, and outputs a browsable `index.html` gallery. Use when the user asks to generate AI images, create pictures with DALL-E, batch-produce image assets, render AI art, or build an image gallery from text prompts.
Transcribe audio via OpenAI Audio Transcriptions API (Whisper). Use when the user wants to transcribe, convert speech to text, extract words from audio or voice recordings, generate a transcript from an audio file, or perform speech recognition on m4a, ogg, or wav files using the Whisper model.
Local speech-to-text with the Whisper CLI (no API key). Use when the user needs to transcribe audio, convert speech to text, generate subtitles, translate spoken language, or produce SRT/VTT captions from mp3, m4a, or wav files using the local Whisper model.
The agent uses LibAFL, a modular Rust fuzzing library, to build custom fuzzers with fine-grained control over observers, feedback mechanisms, mutators, schedulers, and executors. It supports drop-in libFuzzer replacement mode via libFuzzer.a, fully custom fuzzer construction with InProcessExecutor and coverage-guided feedback, multi-core fuzzing with Launcher, crash deduplication via BacktraceObserver, and dictionary-based token mutations. The agent applies LibAFL when standard fuzzers like libFuzzer or AFL++ lack needed customization — such as custom mutation strategies, novel feedback mechanisms, non-standard target architectures, or fuzzing research requiring component-level control over the fuzzing loop, corpus management, and sanitizer integration.
