مرور مهارت‌ها

Runs CodeQL static analysis for security vulnerability detection using interprocedural data flow and taint tracking. Applicable when finding vulnerabilities, running a security scan, performing a security audit, running CodeQL, building a CodeQL database, selecting query rulesets, creating data extension models, or processing CodeQL SARIF output. NOT for writing custom QL queries or CI/CD pipeline setup.

Browser automation, debugging, and performance analysis using Puppeteer CLI scripts. Use for automating browsers, taking screenshots, analyzing performance, monitoring network traffic, web scraping, form automation, and JavaScript debugging.

Expert bash/shell scripting system across ALL platforms. PROACTIVELY activate for: (1) ANY bash/shell script task, (2) System automation, (3) DevOps/CI/CD scripts, (4) Build/deployment automation, (5) Script review/debugging, (6) Converting commands to scripts. Provides: Google Shell Style Guide compliance, ShellCheck validation, cross-platform compatibility (Linux/macOS/Windows/containers), POSIX compliance, security hardening, error handling, performance optimization, testing with BATS, and production-ready patterns. Ensures professional-grade, secure, portable scripts every time.

Build a Sales Qualification Pack (ICP + disqualification rules, qualification scorecard, discovery/qualification script, CRM note template, and pipeline hygiene rules). Use to fix pipeline quality and stop wasting time on wrong leads. Category: Sales & GTM.

Build real-time conversational AI voice engines using async worker pipelines, streaming transcription, LLM agents, and TTS synthesis with interrupt handling and multi-provider support

Database migration best practices for schema changes, data migrations, rollbacks, and zero-downtime deployments across PostgreSQL, MySQL, and common ORMs (Prisma, Drizzle, Kysely, Django, TypeORM, golang-migrate).

Build, test, inspect, install, and deploy MCP servers with FastMCP in Python. Use when creating a new MCP server, wrapping an API or database as MCP tools, exposing resources or prompts, or preparing a FastMCP server for Claude Code, Cursor, or HTTP deployment.

Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence. Use when asked to "open in browser", "test the site", "take a screenshot", or "dogfood this". (gstack)

Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when CUDA is unavailable. Supports GGUF quantization (1.5-8 bit) for reduced memory and 4-10× speedup vs PyTorch on CPU.

Universal LaTeX document skill: create, compile, and convert any document to professional PDF with PNG previews. Supports resumes, reports, cover letters, invoices, academic papers, theses/dissertations, academic CVs, presentations (Beamer), scientific posters, formal letters, exams/quizzes, books, cheat sheets, reference cards, exam formula sheets, fillable PDF forms (hyperref form fields), conditional content (etoolbox toggles), mail merge from CSV/JSON (Jinja2 templates), version diffing (latexdiff), charts (pgfplots + matplotlib), tables (booktabs + CSV import), images (TikZ), Mermaid diagrams, AI-generated images, watermarks, landscape pages, bibliography/citations (BibTeX/biblatex), multi-language/CJK (auto XeLaTeX), algorithms/pseudocode, colored boxes (tcolorbox), SI units (siunitx), Pandoc format conversion (Markdown/DOCX/HTML ↔ LaTeX), and PDF-to-LaTeX conversion of handwritten or printed documents (math, business, legal, general). Compile script supports pdflatex, xelatex, lualatex with auto-detection, latexmk backend, texfot log filtering, PDF/A output, and verbosity control (--verbose/--quiet). Empirically optimized scaling: single agent 1-10 pages, split 11-20, batch-7 pipeline 21+. Use when user asks to: (1) create a resume/CV/cover letter, (2) write a LaTeX document, (3) create PDF with tables/charts/images, (4) compile a .tex file, (5) make a report/invoice/presentation, (6) anything involving LaTeX or pdflatex, (7) convert/OCR a PDF to LaTeX, (8) convert handwritten notes, (9) create charts/graphs/diagrams, (10) create slides, (11) write a thesis or dissertation, (12) create an academic CV, (13) create a poster, (14) create an exam/quiz, (15) create a book, (16) convert between document formats (Markdown, DOCX, HTML to/from LaTeX), (17) generate Mermaid diagrams for LaTeX, (18) create a formal business letter, (19) create a cheat sheet or reference card, (20) create an exam formula sheet or crib sheet, (21) condense lecture notes/PDFs into a cheat sheet, (22) create a fillable PDF form with text fields/checkboxes/dropdowns, (23) create a document with conditional content/toggles (show/hide sections), (24) generate batch/mail-merge documents from CSV/JSON data, (25) create a version diff PDF (latexdiff) highlighting changes between documents, (26) create a homework or assignment submission with problems and solutions, (27) create a lab report with data tables, graphs, and error analysis, (28) encrypt or password-protect a PDF, (29) merge multiple PDFs into one, (30) optimize/compress a PDF for web or email, (31) lint or check a LaTeX document for common issues, (32) count words in a LaTeX document, (33) analyze document statistics (figures, tables, citations), (34) fetch BibTeX from a DOI, (35) convert a Graphviz .dot file to PDF/PNG, (36) convert a PlantUML .puml file to PDF/PNG, (37) create a one-pager/fact sheet/executive summary, (38) create a datasheet or product specification sheet, (39) extract pages from a PDF (page ranges, odd/even), (40) check LaTeX package availability before compiling, (41) analyze citations and cross-reference with .bib files, (42) debug LaTeX compilation errors, (43) make a document accessible (PDF/A, tagged PDF), (44) create lecture notes or course handouts, (45) fill an existing PDF form (fillable fields or non-fillable with annotations), (46) extract text or tables from a PDF (pdfplumber, pypdf), (47) OCR a scanned PDF to text (pytesseract), (48) create a PDF programmatically with reportlab (Canvas, Platypus), (49) rotate or crop PDF pages (pypdf), (50) add a watermark to an existing PDF, (51) extract metadata from a PDF (title, author, subject).

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.

This skill details how to conduct cloud security audits using Center for Internet Security benchmarks for AWS, Azure, and GCP. It covers interpreting CIS Foundations Benchmark controls, running automated assessments with tools like Prowler and ScoutSuite, remediating failed controls, and maintaining continuous compliance monitoring against CIS v5 for AWS, v4 for Azure, and v4 for GCP.

Unified team skill for interactive component crafting. Vanilla JS + CSS interactive components with zero dependencies. Research -> interaction design -> build -> a11y test. Uses team-worker agent architecture with roles/ for domain logic. Coordinator orchestrates pipeline with GC loops and sync points. Triggers on "team interactive craft", "interactive component".

Azure Cosmos DB JavaScript/TypeScript SDK (@azure/cosmos) for data plane operations. Use for CRUD operations on documents, queries, bulk operations, and container management. Triggers: "Cosmos DB", "@azure/cosmos", "CosmosClient", "document CRUD", "NoSQL queries", "bulk operations", "partition key", "container.items".

Azure AI Agents Persistent SDK for .NET. Low-level SDK for creating and managing AI agents with threads, messages, runs, and tools. Use for agent CRUD, conversation threads, streaming responses, function calling, file search, and code interpreter. Triggers: "PersistentAgentsClient", "persistent agents", "agent threads", "agent runs", "streaming agents", "function calling agents .NET".

Diagnoses and fixes CI/CD pipeline failures. Use when user mentions 'CI', 'GitHub Actions', 'GitLab CI', 'ビルドエラー', 'テスト失敗', 'パイプライン', 'CIが落ちた', or asks to analyze build/test failures. Do NOT load for: ローカルビルド, 通常の実装作業, レビュー, セットアップ.

Public-records OSINT investigation framework — SEC EDGAR filings, USAspending contracts, Senate lobbying, OFAC sanctions, ICIJ offshore leaks, NYC property records (ACRIS), OpenCorporates registries, CourtListener court records, Wayback Machine archives, Wikipedia + Wikidata, GDELT news monitoring. Entity resolution across sources, cross-link analysis, timing correlation, evidence chains. Python stdlib only.

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.

Chief Security Officer mode. Infrastructure-first security audit: secrets archaeology, dependency supply chain, CI/CD pipeline security, LLM/AI security, skill supply chain scanning, plus OWASP Top 10, STRIDE threat modeling, and active verification. Two modes: daily (zero-noise, 8/10 confidence gate) and comprehensive (monthly deep scan, 2/10 bar). Trend tracking across audit runs. Use when: "security audit", "threat model", "pentest review", "OWASP", "CSO review". (gstack) Voice triggers (speech-to-text aliases): "see-so", "see so", "security review", "security check", "vulnerability scan", "run security".

GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.