Claude Skills by yanacuti1121

Research prediction markets as data sources or oracle signals for products, agents, dashboards, and corporate decision intelligence. Use for source-grounded analysis of market-implied probabilities, caveats, and integration patterns without investment advice.

Review prediction-market, basket, oracle, and trading-agent workflows for compliance, safety, data-quality, privacy, and execution risk. Use before any workflow handles venue auth, user portfolio data, API keys, or trade planning.

The Common Vulnerability Scoring System (CVSS) is the industry standard

Prisma ORM patterns for TypeScript backends — schema design, query optimization, transactions, pagination, and critical traps like updateMany returning count not records, $transaction timeouts, migrate dev resetting the DB, @updatedAt skipped on bulk writes, and serverless connection exhaustion.

Nanosecond-speed process existence check by PID. sindresorhus/process-exists patterns, cross-platform PID validation, zombie detection, and watchdog loop for monitoring long-running agent sandboxes. Sources: sindresorhus/process-exists.

System process list monitoring and PID management for agent sandboxes. Cross-platform process enumeration, CPU/memory per-process metrics, filtering by name/PID, and detecting unexpected child spawns. Sources: sindresorhus/ps-list.

'Processes STIX 2.1 threat intelligence bundles delivered via TAXII 2.1

Use when designing and simulating a simple CPU/processor from first principles — instruction set design, fetch-decode-execute cycle, a minimal assembler. Not for general low-level programming or existing CPU architecture study. Triggers on: 'design a CPU from scratch', 'build a simple processor', 'implement an instruction set', 'fetch decode execute cycle', 'write a toy CPU simulator', 'build my own assembler'. Covers ISA design, the FDE cycle, register/ALU design, and software-simulated vs H...

Code profiling and benchmarking laws from 5 repos. Statistical benchmark harnesses, micro-second delta timing, string-processing performance comparison, IOPS throttle monitoring, and bundle size analysis. Sources: bestiejs/benchmark.js, alexhuszagh/rust-benchmarks, sindresorhus/time-span, holepunchto/hypercore, hughsk/disc.

'Develops comprehensive threat actor profiles for APT groups, criminal

Prometheus metrics for LLM/AI inference telemetry. Token throughput counters, KV-cache hit/miss rates, latency histograms, model queue depth gauges, and Grafana dashboard patterns. Sources: prometheus/client_golang (Apache-2.0).

Prometheus scrape configuration, recording rules, alerting rules, and Alertmanager routing. Service discovery, relabeling, federation, and cluster-wide agent health monitoring. Sources: prometheus/prometheus (Apache-2.0).

Minimize token cost and latency with Anthropic prompt caching — cache breakpoints, static vs dynamic block separation, skills snapshot pre-computation, context invalidation strategy, and 90% cost reduction patterns for repeated system content. Use when asked about "prompt caching", "cache breakpoint", "reduce token cost", "Anthropic caching", "cache_control", "reuse system prompt", "skills snapshot", "context invalidation", "token budget", "90% cache hit", "cost optimization for Claude", or "...

Design, version, and optimize LLM prompts — few-shot examples, chain-of-thought, role framing, constraint injection, A/B testing, and prompt version management. Use when asked to "write a system prompt", "improve this prompt", "prompt isn't working", "add few-shot examples", "prompt A/B test", or "how do I get better outputs from the model". Do NOT use for: RAG retrieval design — use `rag-architect`. UI for streaming responses — use `llm-ui-patterns`.

Build cognitive security firewalls against prompt injection, jailbreak attacks, and Unicode homoglyph smuggling. Instruction-data segregation, jailbreak token filtering, LLM-as-a-Judge semantic guard, and system prompt reinforcement.

Protocol Buffers binary serialization for inter-agent messaging. .proto schema definition, encode/decode, backward-compatible field evolution, and performance benchmarks vs JSON. Sources: protocolbuffers/protobuf (BSD-3-Clause).

protobuf.js dynamic .proto parsing and code generation for Node.js. Runtime schema loading, static code generation, reflection API, and JSON interop for agent message contracts. Sources: protobufjs/protobuf.js (BSD-3-Clause).

Synthesize red team findings and blue team fixes into a structured security report. Use when the user wants a consolidated security report after red-team-check and blue-team-fix have been run, or when presenting security findings to stakeholders. Produces: executive summary, finding table with evidence, fix status, and open items.

Build type-safe AI agents with PydanticAI — define Agent with result_type, system_prompt, tools (function tools + structured tools), deps injection, and run sync/async with full Pydantic validation on inputs and outputs.

Facebook Pyre static type analysis patterns for Python. Incremental type checking, taint tracking for data-flow security analysis, pysa rules, and integrating Pyre into CI pipelines. Sources: facebook/pyre-check (MIT).

Qdrant vector database — collections, upsert, search, filtering, payloads, sparse vectors, BM25

Quarkus 3.x LTS architecture patterns with Camel for messaging, RESTful API design, CDI services, data access with Panache, and async processing. Use for Java Quarkus backend work with event-driven architectures.

Quarkus Security best practices for authentication, authorization, JWT/OIDC, RBAC, input validation, CSRF, secrets management, and dependency security.

Test-driven development for Quarkus 3.x LTS using JUnit 5, Mockito, REST Assured, Camel testing, and JaCoCo. Use when adding features, fixing bugs, or refactoring event-driven services.

Verification loop for Quarkus projects: build, static analysis, tests with coverage, security scans, native compilation, and diff review before release or PR.

Raft distributed consensus algorithm patterns. Leader election, log replication, safety invariants, snapshot/compaction, and membership changes for replicated agent state machines. Sources: hashicorp/raft (MPL-2.0).

Node.js-native Raft consensus implementation for embedding distributed consensus directly into agent processes. In-process leader election, log replication via TCP, and state machine callbacks without external services. Sources: skiff-project/skiff (MIT).

Design and audit Retrieval-Augmented Generation (RAG) pipelines — chunking strategy, embedding choice, hybrid search, retrieval evaluation, and failure diagnosis. Use when asked to "build a RAG system", "fix retrieval quality", "improve RAG accuracy", "chunk documents", "add semantic search", or when an LLM keeps hallucinating despite having access to docs. Do NOT use for: fine-tuning decisions — that is a separate trade-off.

Evaluate RAG pipelines with Ragas — measure faithfulness, answer relevancy, context precision/recall, and noise sensitivity using LLM-as-judge metrics; run automated test suite generation with TestsetGenerator; integrate with LangChain, LlamaIndex, and CI pipelines.

RAM-backed virtual disk patterns for zero-latency sandbox scratch space. Linux tmpfs mounts, Node.js ramdisk creation, size-capped temp volumes, auto-cleanup on process exit, and race-condition-free temp directory isolation. Sources: mafintosh/ramdisk, Linux tmpfs docs.

Raw IP packet analysis for agent network diagnostics. ICMP ping, TCP port probing, packet crafting, and network path tracing using raw sockets — for Sandbox network isolation validation. Sources: indutny/raw-socket.

ReAct (Reason+Act) agent loop — the canonical Observe/Think/Act cycle. Tool registry, stop conditions, turn budget, observation formatting. Every 2026 agent framework runs this loop under the hood. Sources: rohitg00/ai-engineering-from-scratch (Apache-2.0).

Diagnose and fix React component health issues — unnecessary re-renders, missing keys, stale closures, useEffect misuse, memory leaks, bundle size regressions, and dangerouslySetInnerHTML XSS risks. Use when asked to "React performance audit", "why does this re-render", "React health check", "stale closure", "missing dependency array", "React memory leak", "React XSS", "fix useEffect", "component re-renders too often", "React bundle bloat", or "React anti-patterns". Do NOT use for: full E2E t...

React 18/19 hooks discipline, server/client component boundaries, Suspense + error boundaries, form actions, state management, and accessibility-first composition. Use when building or reviewing React/Next.js components.

React and Next.js performance optimization — eliminate waterfalls, reduce bundle size, optimize renders, server-side patterns. 70+ rules by impact priority. Profile first, optimize second.

React component testing with React Testing Library + Vitest/Jest + MSW network mocking + axe accessibility assertions. Covers unit, integration, and component vs E2E decision boundary.

Read-only bind mount patterns for safe source code analysis. Mount workspace as ro inside containers, overlayfs for copy-on-write analysis, Docker ro bind mounts, detecting unauthorized writes before they reach source. Sources: docker/cli, OCI spec, overlayfs docs.

Use when you receive review feedback on your own code or changes. Triggers on: 'reviewer said', 'PR feedback', 'review comments', 'address the feedback', 'fix the review'. Classifies feedback into must-fix / optional / reject-with-reason.

OSINT graph investigation platform — visual, flexible, extensible. Dùng khi cần map relationships, investigate entities, trace connections giữa người/tổ chức/domains.

Recover deleted files from disk images and storage media using PhotoRec's

'Executes structured recovery from a ransomware incident following NIST

Design composable recommendation, ranking, and feed pipelines using the six-stage Source→Hydrator→Filter→Scorer→Selector→SideEffect framework popularized by xAI's open-sourced For You algorithm. Use this skill whenever the user is building any system that picks "the top K items for a (user, context)" — social feeds, content CMSs, RAG rerankers, task prioritizers, notification triage, search reranking, ad ranking.

Use when the user asks for repeated rollouts, marked decision processes, high-dimensional search, stochastic optimization, local-optima exploration, ensemble comparison, or recursive reasoning with a visible evidence trail.

Offensive security review of the current local codebase — scan for OWASP Top 10 vulnerabilities, hardcoded secrets, misconfigurations, and broken access control. Use when the user asks to find security issues, audit the codebase for vulnerabilities, run a red team check, or scan for security bugs. Requires ownership confirmation via security-scope-gate before starting. Only runs against repos the user owns.

Upgrades existing websites and apps to premium quality. Audits current design, identifies generic AI patterns, and applies high-end design standards without breaking functionality. Works with any CSS framework or vanilla CSS.

Redis data structure patterns, caching strategies, distributed locks, rate limiting, pub/sub, and connection management for production applications.

Apply systematic refactoring patterns — strangler fig for legacy migration, seam detection, extract method/class, introduce parameter object, replace conditional with polymorphism, and safe refactoring sequencing. Use when asked to "refactor this", "clean up legacy code", "this is unmaintainable", "too many if statements", "extract this logic", or "how do I migrate without breaking everything". Do NOT use for: full rewrites — that is an architecture decision, not a refactoring task.

Runtime metadata reflection for TypeScript dependency injection. Decorator-based class metadata, design:type/paramtypes annotation, custom metadata keys, and DI container bootstrap. Sources: rbuckton/reflect-metadata (Apache-2.0).

Reflexion verbal reinforcement learning — Actor/Evaluator/Self-Reflector loop with episodic memory. No gradient updates. ALFWorld +X% over ReAct, HumanEval SOTA at publish time. Foundation for CLAUDE.md learnings, Letta sleep-time compute, learn-rule. Sources: rohitg00/ai-engineering-from-scratch (Apache-2.0).

Prevent ReDoS (Regex Denial of Service) by escaping user-supplied regex special characters. escape-string-regexp patterns, catastrophic backtracking detection, safe dynamic regex construction, and timeout guards. Sources: sindresorhus/escape-string-regexp.