向量嵌入

Mindflow

hrshitkunwar-tech

Collects saved content and bookmarks from social media platforms, organizes them using AI, and generates searchable mind maps for fast knowledge discovery and insight generation.

SKILL.md — Web Research

iamthetonyb

Never present speculation as fact

numpy-sorting

cuba6112

Sorting and searching algorithms including O(n) partitioning, binary search, and hierarchical multi-key sorting. Triggers: sort, argsort, partition, searchsorted, lexsort, nan sort order.

ace-search

PandAddKai

'Use when the task is read-only codebase discovery/search (查找/搜索/定位): find files, functions, classes, call sites, tests, or "where is X implemented" questions; prefer Augment Context Engine MCP (augment-context-engine) codebase retrieval over grep/rg.'

llm-tldr

Tiberriver256

Summarize and navigate this repo with the llm-tldr CLI (tree/structure/context/semantic search). Use when you need LLM-ready summaries of files or symbols, fast codebase orientation, or behavior-based search before editing.

local-memory-search

blackbasilisk

Local, offline semantic search over OpenClaw memory files (MEMORY.md and memory/*.md) using Python embeddings + FAISS (no online LLM/API). Use when Dave asks for "memory search" without cloud billing, wants semantic recall across notes, or wants to share a reusable local-memory semantic search skill.

zvec-grep-compare-not-clone

IgorGanapolsky

zg (zvec-grep) is a local-first search layer (ripgrep + BM25 + vector), not a ThumbGate clone. Steal the four-route FORMAT onto existing rails; never install @zvec/zvec-grep. Slash: /zvec-grep-compare-not-clone.

bulk-remove-classname

yuichiroharai

Removes a specified class name from HTML/JSX files using the classname-search CLI tool. Use when you need to safely remove all occurrences of a specified class name from target files.

sandbox

TakaGoto

"Spin up a minimal RAG pipeline instantly to experiment with"

conduct-ad-research

feed-mob

Conduct comprehensive advertising research based on confirmed campaign parameters. Use web search tools (DuckDuckGo or Tavily) to gather insights on target audiences, platform trends, competitor strategies, industry benchmarks, and creative best practices. Generate a structured campaign report with actionable recommendations for ad strategy development.

building-native-ui

Jackiexiao

（中文）Complete guide for building beautiful apps with Expo Router. Covers fundamentals, styling, components, navigation, animations, patterns, and native tabs.

benchmark

TakaGoto

"Benchmark your RAG pipeline's performance"

media-transcript-search

joshuaroll

Extract and search transcripts from multimedia sources (primarily YouTube). Allows finding specific topics within long content without watching the whole video.

semantic-search

amoscicki

Semantic codebase search with pgvector. Finds code by MEANING. USE THIS FIRST before grep/glob. Returns file paths + byte ranges for precise reads. Multi-project support with auto-discovery — no manual setup needed.

find-skills

vilin1927

Search and discover Claude Code skills from the community. Helps users find, evaluate, and install skills from GitHub repositories and the skills registry.

skill-holded

fermatrix

"Access Holded ERP via REST API. Use for contacts, invoices, estimates, sales orders, purchase orders, expenses, credit notes, proformas, waybills, products, warehouses, accounting accounts, ledger, taxes, and treasury. Supports searching, creating, updating, paying, sending and downloading PDFs for all document types. Triggers: Holded de ENZO, busca contacto en Holded, factura Holded, pedido Holded, producto Holded, contabilidad Holded, tesorería Holded."

UI/UX Pro Max - Design Intelligence

zwmmm

roadmap

TakaGoto

"View your learning progress and recommended next steps"

agent-context-partitioner

curiositech

Algorithmic skill for designing evidence-backed partitions of LLM context or task DAGs across K agents. Covers two modes: (A) online/reactive spawning under context pressure, and (B) static pre-planned DAG execution with graph min-cut optimization. Encodes algorithms for sequence-stable clustering, dynamic K selection, token-budget-aware partitioning, and handoff protocol design. Use when designing multi-agent systems, deciding when and how many agents to spawn, or optimizing a task DAG for token efficiency. NOT for one bounded task that does not benefit from parallelism, runtime scheduling or authorization policy, or comparing vectors whose embedding-space identities are unknown or incompatible.

deep-search-research

canxia-hub

Research-agent workflow for deep multi-source search, platform-aware routing, source normalization, evidence tracking, and structured report delivery. Use when the user asks to perform or build deep research across the web or multiple platforms; when a task needs a research plan before searching; when results must include citations, source credibility, or cross-platform synthesis; or when building/iterating a deep-search skill, adapters, retrieval pipelines, or research reports.

firecrawl

YPYT1

Web search and scraping via Firecrawl API. Use when you need to search the web, scrape websites (including JS-heavy pages), crawl entire sites, or extract structured data from web pages. Requires FIRECRAWL_API_KEY environment variable.

chroma

majiayu000

Open-source embedding database for AI applications. Store embeddings and metadata, perform vector and full-text search, filter by metadata. Simple 4-function API. Scales from notebooks to production clusters. Use for semantic search, RAG applications, or document retrieval. Best for local development and open-source projects.

fd-search

wizact

"REQUIRED for file searches. Use instead of bash find - faster, simpler syntax, regex support, respects .gitignore. Syntax: fd [pattern] [path]. Common: fd -e ext, fd -t f/d, fd -g '*.txt', fd -S +100m, fd --changed-within 1day"

qavr-memory

kimasplund

Q-Value Augmented Vector Retrieval for learned memory ranking. Tracks which memories are most useful over time and prioritizes them in retrieval.