Agent Skills by 张烨
Showing 145 vetted skills indexed across 2 GitHub repositories.
superfhpgoogle_meet
Join Google Meet calls, transcribe live captions, and speak via OpenAI Realtime.
superfhpparallel-cli
Automate web searches, deep research, and data enrichment with Parallel CLI.
superfhpgitnexus-explorer
Index codebases into interactive knowledge graphs with a web UI.
superfhpqmd
Index and query markdown notes with BM25, vector search, and LLM reranking.
superfhpdrug-discovery
Search ChEMBL for bioactive compounds and calculate drug-likeness scores.
superfhpduckduckgo-search
Search the web via DuckDuckGo for text, news, images, and video results.
superfhpscrapling
Extract website data using HTTP requests, stealth browser automation, and spider crawling.
superfhpbioinformatics
Access bioinformatics skills for genomic, transcriptomic, and metagenomic analysis.
superfhpdomain-intel
Perform passive domain reconnaissance using Python stdlib for subdomains, SSL, WHOIS, and DNS.
superfhpdocker-management
Automate Docker container, image, and network lifecycle management.
superfhpinference-sh-cli
Runs AI apps from the command line via the infsh CLI.
superfhpsherlock
Identify social media accounts for a username across 400+ platforms.
superfhposs-forensics
Investigate GitHub repositories for supply chain attacks and compromised code.
superfhp1password
Manage secrets in 1Password using the 1Password CLI.
superfhpfitness-nutrition
Plan workouts and track nutrition using wger and USDA FoodData Central APIs.
superfhpneuroskill-bci
Read real-time brain and body metrics from BCI wearables into AI interactions.
superfhpnemo-curator
Curate text, image, video, and audio data for LLM training datasets.
superfhphuggingface-accelerate
Unify distributed training libraries for PyTorch with automatic device placement and mixed precision.
superfhpchroma
Vectorize text and manage a vector database for semantic search.
superfhpsimpo-training
Align language models by optimizing preferences with SimPO without a reference model.
superfhpslime-rl-training
Automate RL training with Megatron-LM and SGLang rollout generation.
superfhpguidance
Generate valid JSON, XML, and code using regex and grammatical constraints.
superfhpclip
Classify images, search semantically, and moderate content with OpenAI's CLIP model.
superfhplambda-labs-gpu-cloud
Launch and manage Lambda Labs GPU cloud instances for machine learning workloads.
Frequently Asked Questions About 张烨
FAQPage SchemaWhat tasks can I accomplish with superfhp's skill registry?▼
You can fine-tune LLMs with LoRA/QLoRA/DPO, serve models via vLLM or llama.cpp, run semantic vector search with FAISS, Qdrant, or Pinecone, automate GitHub PRs and code review, debug Python and Node.js, and generate images, music, diagrams, and videos from text prompts.
Who is the target audience for these skills?▼
ML engineers running distributed training with PyTorch FSDP or Megatron-LM, researchers writing NeurIPS/ICML papers, developers automating GitHub and Linear issue workflows, and creative coders building TouchDesigner, p5.js, or Manim visualizations on macOS and Linux.
How do I install and run these skills?▼
Skills run inside the Hermes Agent framework as SKILL.md files invoked from the terminal. Many require prerequisite CLI binaries such as gh, xurl, himalaya, or remindctl, plus environment variables like LINEAR_API_KEY, NOTION_API_KEY, or AIRTABLE_API_KEY for platform integrations.
Are these skills open source and free to use?▼
Nearly all skills carry MIT licenses from authors including Hermes Agent, Orchestra Research, and community contributors. One exception is the powerpoint skill, which is proprietary with terms defined in its bundled LICENSE.txt file.
What dependencies do the machine learning skills require?▼
Training skills depend on torch, transformers, peft, trl, datasets, and accelerate; serving skills need vllm or llama-cpp-python; evaluation uses lm-eval. GPU skills target NVIDIA CUDA, with llama.cpp additionally supporting Apple Silicon, ROCm, and Intel GPUs.