NLP

Use before creating new stories or when user asks "check for conflicts", "find duplicates", "review story overlaps", or "detect story conflicts" - scans story-tree database to identify duplicate stories, scope overlaps, and competing approaches using lightweight text similarity (no external dependencies).

ユーザープロンプトを分析し、topic_type（instruction/question/context）を判定する。

This skill should be used when the user asks to "list posters", "get poster info", "fetch context document", "get latest poster", "check poster count", "show poster details", "get context for poster", "show enrichment quality", "diagnose enrichment", "show pulse script", "get all poster data", "check what data we have", or mentions Supabase, posters, context documents, enrichment, or database queries. Provides commands and scripts for querying Local Pulse poster data from Supabase.

Branch the current Claude Code session into a new tmux session using --resume. Creates a parallel conversation that can diverge independently. Use when you want to explore an alternative approach without losing your current context.

Define boundaries where claims or recommendations apply versus break down. Use when: (1) asked to define scope or boundaries, (2) claim presented as universal truth, (3) explaining why pattern works here but not there, (4) advice given without specifying applicable context.

Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.

Detect and mask PII (names, emails, phones, SSN, addresses) in text and CSV files. Multiple masking strategies with reversible tokenization option.

Enhance an existing PR by analyzing the diff against the description and updating with examples, diagrams, and context

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Sử dụng khi: tạo MCP server, xây dựng tool server.

Enterprise context and session management with token budget optimization and state persistence

[Aesthetics] Extracts recurring visual patterns from references: contrast habits, shapes, density/whitespace balance, rhythm, textures. Produces Aesthetic Pattern Library as Data-Sheet nodes in Brain canvas.

Synchronize task state and metadata across Claude, Cursor, and Factory Droid sessions. Use when handing off tasks between platforms, sharing plans, or updating external trackers like Linear or Jira.

Loading and using pretrained models with Hugging Face Transformers. Use when working with pretrained models from the Hub, running inference with Pipeline API, fine-tuning models with Trainer, or handling text, vision, audio, and multimodal tasks.

Systematic verification of codebase context before code generation to prevent assumption bugs. Trigger keywords: context, assumptions, helpers, authentication, current_user, verify, validate context

Fetch source language (Greek/Hebrew) word data for Bible verses. Use this when users want to study Greek or Hebrew words, understand original language meanings, or analyze source text morphology. The skill retrieves Macula source language data and merges it with Strong's dictionary entries to provide comprehensive linguistic information. Also supports direct Strong's number lookups and English word searches across all Greek/Hebrew lexicon entries.

Automated literature search using the `lit` CLI tool for comprehensive paper discovery, citation analysis, and bibliography management.Use this skill when:- Conducting systematic literature reviews- Building citation networks from seed papers- Searching for academic papers on a topic- Fetching and managing BibTeX references- Downloading and extracting text from PDFs- Managing research projects with paper collectionsPrerequisites:- BRIGHTDATA_API_KEY for Google Scholar searches- UNPAYWALL_EMAIL for Open Access PDF lookup (optional)- OPENALEX_MAILTO for higher API rate limits (optional)

Extract topics from text collections using LDA (Latent Dirichlet Allocation) with keyword extraction and topic visualization.

Structure for creating hands-on exercises in the Physical AI textbook.

Inspect GitHub PR checks with gh, pull failing GitHub Actions logs, summarize failure context, then create a fix plan and implement after user approval. Use when a user asks to debug or fix failing PR CI/CD checks on GitHub Actions and wants a plan + code changes; for external checks (e.g., Buildkite), only report the details URL and mark them out of scope.

Foundation skill for integrating OpenAI ChatKit framework with custom backends. This skill should be used for initial ChatKit setup including server implementation, React component integration, authentication, context injection, and database persistence. For streaming UI patterns use chatkit-streaming. For interactive widgets and actions use chatkit-actions.