Claude Skills by yanacuti1121

Establish SAML 2.0 identity federation between on-premises Active Directory

'Builds comprehensive identity governance and lifecycle management processes

'Builds real-time incident response dashboards in Splunk, Elastic, or

'Designs and documents structured incident response playbooks that define

Build collaborative forensic incident timelines using Timesketch to ingest,

Build an automated pipeline to defang indicators of compromise (URLs,

OpenCTI is an open-source platform for managing cyber threat intelligence

Build structured communication templates for malware incidents including

Establish a structured operational process to triage, test, and deploy

Implement a phishing report button in email clients with automated triage

'Builds a structured ransomware incident response playbook aligned with

Deploy and configure the Havoc C2 framework with teamserver, HTTPS listeners,

Apply bottom-up and top-down role mining techniques to discover optimal

Build a structured SOC escalation matrix defining severity tiers, response

'Builds SOC performance metrics and KPI tracking dashboards measuring

'Builds a structured SOC incident response playbook for ransomware attacks

Build comprehensive threat actor profiles using open-source intelligence

Deploy MISP (Malware Information Sharing Platform) to aggregate, correlate,

Build a systematic threat hunt hypothesis framework that transforms threat

Build automated threat intelligence enrichment pipelines in Splunk Enterprise

'Builds automated threat intelligence feed integration pipelines connecting

Building a Threat Intelligence Platform (TIP) involves deploying and

Implement a vulnerability aging dashboard and SLA tracking system to

Deploy DefectDojo as a centralized vulnerability management dashboard

Build a vulnerability exception and risk acceptance tracking system with

'Builds a structured vulnerability scanning workflow using tools like

Bridge Bunyan structured JSON logs to syslog and centralized log aggregators. bunyan-syslog stream, log level mapping, RFC5424 structured data, and dual-write (file + syslog) audit patterns. Sources: trentm/node-bunyan-syslog.

Discovering and accessing unprotected pages, APIs, and administrative

Production caching patterns and memory efficiency laws from 5 repos. In-memory TTL caches, unified KV interfaces, LRU eviction math, consistent hashing for distributed load, and GC thrash prevention via object pooling. Sources: node-cache/node-cache, jaredwray/keyv, isaacs/node-lru-cache, memcached/memcached, tweenjs/tween.js.

Design and implement caching — strategy selection (cache-aside, read-through, write-through, write-behind), cache invalidation, TTL tuning, Redis patterns, cache stampede prevention, and distributed cache consistency. Use when asked to "add caching", "Redis cache", "cache invalidation", "reduce database load", "cache-aside", "CDN caching", "stale-while-revalidate", "cache stampede", "cache warming", or "response is too slow and hits the DB every time". Do NOT use for: HTTP browser caching hea...

cAdvisor container resource metrics from Linux cgroups. CPU throttling detection, OOM kill monitoring, memory working set vs cache, network I/O per container, and integration with Prometheus scraping. Sources: google/cadvisor (Apache-2.0).

omp — AI coding agent viết bằng Rust (27K lines), hash-anchored edits (-61% output tokens), LSP+debugger tích hợp, 32 tools, 40+ LLM providers. Fork của Pi platform.

Linux capability management for agent processes. Drop unnecessary capabilities before exec, audit current capability sets, block CAP_SYS_ADMIN/CAP_NET_ADMIN/CAP_SETUID, and implement least-privilege subprocess execution. Sources: genuinetools/bane (renamed from bansh), libcap.

Query the Yana AI codebase knowledge graph via codebase-memory-mcp. Use instead of grep/glob when exploring call chains, finding callers/callees, understanding architecture, or tracing impact of changes. Triggers on: 'who calls X', 'trace path', 'find callers', 'search graph', 'cbm', 'knowledge graph', 'what calls', 'call chain', 'what uses', 'where is X defined', 'architecture overview', 'impact of changing'.

Danh sách free LLM API providers có rate limits cụ thể — OpenRouter, Groq, Cerebras, Google AI Studio, GitHub Models, Fireworks... Reference khi cần backup API.

Đọc Gmail, báo cáo thư chưa đọc. Dùng khi muốn check inbox nhanh mà không cần mở browser.

Accessibility debugging via Chrome DevTools MCP. Lighthouse a11y audit, accessibility tree (take_snapshot), heading hierarchy, ARIA labels, focus/keyboard nav, color contrast, tap targets. WCAG 2.1 AA. Sources: ChromeDevTools/chrome-devtools-mcp (Apache-2.0).

Automate browser input, navigation, and interaction via Chrome DevTools MCP (Puppeteer-backed). Click, type, fill forms, drag, upload, navigate pages, wait for conditions, emulate devices.

Install, manage, reload, and trigger Chrome extensions via Chrome DevTools MCP. Enables agent-driven extension testing, side-loading unpacked extensions, and extension action automation.

LCP (Largest Contentful Paint) debugging and optimization via Chrome DevTools MCP. Four subparts (TTFB/load-delay/load-duration/render-delay), performance trace workflow, insight analysis, optimization checklist. Good LCP ≤2.5s. Sources: ChromeDevTools/chrome-devtools-mcp (Apache-2.0).

Chrome DevTools MCP server — agent-controlled browser via MCP. take_snapshot (accessibility tree + uids), click/fill/navigate automation, screenshot, evaluate_script, console/network inspection. 41k+ stars. Sources: ChromeDevTools/chrome-devtools-mcp (Apache-2.0).

JavaScript/Node.js memory leak debugging via Chrome DevTools MCP. Heap snapshot workflow (baseline→target→final), memlab analysis, common leak patterns (detached DOM, closures, unbounded caches, event listeners). Sources: ChromeDevTools/chrome-devtools-mcp (Apache-2.0).

Inspect, list, and analyze live browser network requests via Chrome DevTools MCP. Identify slow requests, failed fetches, API payloads, CORS errors, and waterfall bottlenecks.

Record and analyze Chrome performance traces via DevTools MCP. Capture CPU flame graphs, long tasks, layout thrash, and JS execution timelines. Extract actionable insights from trace data.

Capture screenshots and record live screencasts from Chrome via DevTools MCP. Use for visual regression, agent state verification, bug reproduction capture, and frame-by-frame UI analysis.

Design CI/CD pipelines and container builds — Dockerfile best practices, GitHub Actions pipeline structure, pipeline stages, secrets management in CI, deployment strategies, and pipeline failure triage. Use when asked to "set up CI", "write a Dockerfile", "GitHub Actions pipeline", "deployment strategy", "blue-green deploy", "canary release", or "why is CI slow/failing". Do NOT use for: deploy gate authorization — that is YAMTAM's `deploy-gate.sh`.

Implement circuit breaker patterns, canary deployment gates, hallucination telemetry checkers, and automated rollback for agent skill deployments. Stop cascading failures before they propagate across the swarm.

Extract every verifiable claim from a piece of content, check each against sources, and sort into 4 buckets: correct, unsourced, needs-fixing, misleading. Use when asked to 'kiểm chứng luận điểm', 'audit claims', 'verify this article', 'tách luận điểm', 'check facts in this doc', 'đối chiếu với nguồn', or 'is this content accurate'. Do NOT use for: verifying a single claim before stating it — see /fact-check command. Do NOT use for: code correctness — see verification-engine.

Use when setting up a structured Plan→Work→Review→Release cycle for autonomous agents, integrating Claude Code with Codex/Cursor/OpenCode, enforcing spec-first development gates, or protecting against agent self-modification. Triggers on: 'claude-code-harness', 'harness cycle', 'agent release cycle', 'spec gate', 'plan work review release', 'harness-plan', 'harness-work', 'harness-review', 'harness-release', 'codex integration', 'cursor integration', 'opencode integration', 'multi-tool agent ...

Orchestrate Claude Code (Anthropic CLI) as a subagent. Two modes — print mode (one-shot, fast) and PTY/tmux mode (interactive, multi-turn). Use when delegating a coding task to a fresh Claude Code session with restricted tools.