AI & ML

bowen31337 Skill

Scans the codebase for async/await patterns, thread-safety mechanisms, and pool usage, then generates framework-specific concurrency rules and code snippets matched to the detected conventions. Detects: anyio, asyncio (stdlib), threading, concurrent.futures, aiohttp, httpx. Flags anti-patterns: blocking calls (time.sleep, requests.get) inside async functions, missing in-memory locks for shared mutable state, bare thread spawns without join, deprecated asyncio.get_event_loop(). Generates ready-to-paste code snippets for: anyio TaskGroup, asyncio.TaskGroup, asyncio.Lock / anyio.Lock, ThreadPoolExecutor with correct sizing, run_sync / run_in_executor for bridging sync code, fcntl.LOCK_EX for cross-process file safety. Use when: (1) starting work in an async codebase and need to know the team's conventions, (2) reviewing a PR that adds concurrency primitives, (3) a linter or code-review agent needs concrete async/thread-safety rules, (4) debugging a suspected race condition or blocking-in-async bug, (5) on-boardi

jiunbae Bundle

Run bounded parallel planning with independent planner personas and optional multi-LLM workers, then synthesize a single plan via stance-steered consensus with explicit conflicts and open questions. Use only when the user explicitly requests background, parallel, or multi-agent planning, such as "백그라운드 기획", "bg plan", "병렬 기획", "멀티 AI 기획", or "N명이 기획". Do not trigger for ordinary planning or design requests.

jiunbae Bundle

Run bounded parallel code reviews with independent persona reviewers, verify findings adversarially, merge and prioritize by root cause, and optionally repeat a review-fix-verify cycle. Use only when the user explicitly requests a background, parallel, multi-agent, persona, multi-LLM, or iterative review such as "bg review", "멀티 리뷰", "페르소나 리뷰", "리뷰 루프", or "keep reviewing". Do not trigger for an ordinary code review.

veniceai Skill

One or two sentences describing exactly when an agent should load this skill and what it covers. Mention the specific endpoints, parameters, or scenarios so the agent can confidently pick it — vague descriptions hurt skill selection.

veniceai Skill

Call POST /chat/completions on Venice. Covers the OpenAI-compatible request shape, Venice-only venice_parameters (web search, E2EE, characters, thinking control, X search), multimodal inputs (images/audio/video), tool calls, reasoning controls, streaming, prompt caching, structured output, and model feature suffixes.

veniceai Skill

Handle Venice API errors correctly. Covers the StandardError / DetailedError / ContentViolationError / X402InferencePaymentRequired body shapes, every meaningful status code (400, 401, 402, 403, 415, 422, 429, 500, 503, 504), the 402 PAYMENT-REQUIRED header used by x402 inference, 422 content-policy suggested_prompt retry pattern, 429 rate-limit headers, and an exponential-backoff retry strategy with idempotency.

veniceai Skill

Discover Venice models, their capabilities, constraints, and pricing. Covers GET /models (with ?type filter), /models/traits, /models/compatibility_mapping, the ModelResponse schema (capabilities, constraints, pricing per type), and how to use this to pick the right model programmatically.

veniceai Skill

Manage Venice API keys. Covers GET/POST/PATCH/DELETE /api_keys, GET /api_keys/{id}, GET /api_keys/rate_limits, GET /api_keys/rate_limits/log, the two-step /api_keys/generate_web3_key wallet flow, INFERENCE vs ADMIN key types, and per-key consumption limits (USD / DIEM).

veniceai Skill

Discover and use Venice public characters (persona-driven system prompts with a bound model). Covers GET /characters (search/filter/sort), /characters/{slug}, /characters/{slug}/reviews, the Character schema, and how to apply a character via venice_parameters.character_slug in chat completions.

veniceai Skill

Use Venice as a pay-per-call JSON-RPC proxy to 27 EVM, Starknet, and Solana networks. Covers GET /crypto/rpc/networks, POST /crypto/rpc/{network}, chain families and per-family method allowlists, the 1×/2×/4× method-tier pricing model, per-minute + 24-hour credit rate limits, idempotency keys for safe retries, single vs batch requests, and the unsupported stateful/WebSocket methods (eth_subscribe, eth_newFilter, *Subscribe, etc.).

veniceai Skill

Call POST /embeddings on Venice. Covers request shape (input, model, encoding_format, dimensions, user), OpenAI compatibility, response compression (gzip/br), and practical usage for retrieval, clustering, and RAG.

veniceai Skill

Transform existing images with Venice. Covers POST /image/edit (prompt-driven single-image edit), /image/multi-edit (compose multiple images), /image/upscale (2x or 4x upscale), and /image/background-remove. Accepts base64, file upload, or HTTPS URL.

veniceai Skill

Async music / audio-track generation via Venice. Covers the /audio/quote + /audio/queue + /audio/retrieve + /audio/complete lifecycle, lyrics vs instrumental, voice selection, duration, language, speed, model capability probing, and webhook-free polling.

justinwangjp Skill

Agent Teamsを用いたチーム体制を設計・提案するスキル。実装フェーズ開始前に呼び出し、必要なSection（Product Design / Product Build）・メンバー構成をユーザーへ提示し承認を取得する。

justinwangjp Bundle

Comprehensive knowledge of Microsoft Agent Framework for building production AI agents and workflows. Auto-activates for agent building, workflow design, AutoGen migration, and enterprise AI tasks.

jiunbae Bundle

Run bounded parallel implementation with isolated workers (native worktrees or Codex), respect a dependency DAG, and verify each worker's diff before integrating. Use only when the user explicitly requests background, parallel, or multi-agent implementation, such as "백그라운드 구현", "bg impl", or "병렬 구현". Do not trigger for ordinary implementation or code-writing requests.

xd3an Bundle

Reference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST

veniceai Skill

High-level map of the Venice.ai API - base URL, authentication modes, endpoint categories, response headers, pricing model, error shape, and versioning. Load this first when starting any Venice integration.

veniceai Skill

Generate speech from text via POST /audio/speech, and clone a voice via POST /audio/voices. Covers TTS models (Kokoro, Qwen 3, xAI, Inworld, Chatterbox, Orpheus, ElevenLabs Turbo, MiniMax, Gemini Flash, Gradium), voices per family, cloned-voice handles, output formats (mp3/opus/aac/flac/wav/pcm), streaming, prompt/emotion styling, temperature/top_p, and language hints.

veniceai Skill

Generate images with Venice. Covers POST /image/generate (Venice-native), POST /images/generations (OpenAI-compatible), GET /image/styles (style presets), request fields (prompt, dimensions, cfg_scale, seed, variants, style_preset, style_references, aspect_ratio, resolution, safe_mode, watermark), and response formats.

dwmkerr Bundle

Deep research into technical solutions by searching the web, examining GitHub repos, and gathering evidence. Use when the user explicitly says "use the research skill", "use a research agent", or asks for deep/thorough research into implementation options or technologies.

dwmkerr Bundle

This skill should be used when the user asks to "create a skill", "write a skill", "build a skill", or wants to add new capabilities to Claude Code. Use when developing SKILL.md files, organizing skill content, or improving existing skills. Do NOT use for plugin development, hook creation, agent creation, or slash command creation — those have dedicated skills.

powdream Bundle

Cross-review a code change with two independent, strong reviewer agents that check each other, and adversarially verify any single-reviewer finding before adopting it. The two reviewers are codex (via the codex CLI) and opus (via the Agent tool); a finding only one of them raises is sent to the other reviewer, which tries to refute it before it counts. Use this whenever the user asks to review, cross-review, or get a second / independent opinion on a diff, pull request, branch, commit, or set of changed files — for correctness, edge cases, security, error handling, tests, or conventions — especially when they mention "two agents", "cross review", codex, or want findings verified or refuted before they are accepted. Prefer this over a single-agent code review whenever review quality matters or a lone reviewer's false positive / false negative would be costly.

powdream Skill

Use when the user asks you to write a work-instruction, handoff, or continuation prompt for a future/next session to resume in-progress work after context compaction or in a fresh session — triggers like "컴팩션 후/이후 작업 지시 프롬프트 써줘", "이어서 할 작업 프롬프트 만들어줘", "다음 세션 지시문", "핸드오프 프롬프트", "continuation/handoff prompt", or preserving work across a summary boundary.

ikatsuba Bundle

Agent Audit - audits an existing AI agent against the 22 patterns from "Patterns for Building AI Agents" and the 12 factors of production-reliable agents, emitting a scored maturity report with a security gate. Use when reviewing, scoring, or assessing the production-readiness of an AI agent or agent design.

sssemil Skill

Compact the current conversation into a handoff document for another agent to pick up.

sssemil Skill

Stress-test business ideas against the "Reality's Moat" scar tissue framework. Use when a user describes a startup idea, product concept, or business model and wants honest evaluation of its defensibility in a world where AI makes building software free. Also use when comparing multiple ideas, evaluating pivots, or deciding build-vs-buy. Triggers include phrases like "is this idea defensible", "will this get commoditized", "stress test this", "what's the moat", "brutal honesty", or any request to evaluate a business idea's durability.

sssemil Skill

Delegate coding work to OpenCode through safe-opencode. Use when Codex should plan, route the job to a suitable DeepSeek model, review git diff, and validate instead of writing the main code itself.

dwmkerr Bundle

This skill should be used when the user asks to "create evals", "evaluate an agent", "build evaluation suite", or mentions agent testing, graders, or benchmarks. Also suggest when building coding agents, conversational agents, or research agents that need quality assurance.

dwmkerr Bundle

This skill should be used when the user asks to "create an agent team", "run a multi-agent team", "coordinate multiple Claude sessions", "use agent teams", or wants to parallelize work across multiple Claude Code instances.

dwmkerr Bundle

This skill should be used when the user asks to "create a statusline", "customize the status line", "add a custom prompt", or mentions Claude Code statusline. Also suggest when the user wants to display git branch, context usage, model name, or session costs at the bottom of Claude Code.

maragudk Skill

Autonomous experiment loop that iteratively improves a measurable metric. Given a goal, a verify command, and an optional guard, the agent branches, makes one change, measures the result, and keeps or discards the experiment -- repeating indefinitely. Use this skill when the user wants to optimize something measurable through automated experimentation, autonomous improvement loops, or when they mention "autoresearch". Works for any domain with a quantifiable metric (code performance, ML training, build size, test scores, content quality metrics, etc.).

entityprocess Bundle

Run AgentV evaluations and optimize agents through eval-driven iteration. Triggers: run evals, benchmark agents, optimize prompts/skills against evals, compare agent outputs across providers, analyze eval results, offline evaluation of recorded sessions, run autoresearch, optimize unattended, run overnight optimization loop. Not for: writing/editing eval YAML without running (use agentv-eval-writer), analyzing existing traces/JSONL without re-running (use agentv-trace-analyst).

entityprocess Bundle

Author, edit, and lint `governance:` blocks in `*.eval.yaml` files. Use when creating or updating evaluation suites that carry AI-governance metadata (OWASP LLM Top 10, OWASP Agentic Top 10, MITRE ATLAS, EU AI Act, ISO 42001). Also use non-interactively (e.g., from a GitHub Action) to lint changed eval files and report violations against the rules in `references/lint-rules.md`. Do NOT use for running evals or benchmarking — that belongs to agentv-bench.

ikatsuba Bundle

Designs and builds AI agents. Use when working on agent architecture, prompts, model selection, tool design, MCP, memory, context engineering, RAG, graph workflows, multi-agent systems, evaluation, or production reliability.

ikatsuba Skill

Claude Code Setup Audit - reviews CLAUDE.md hierarchy, .claude/settings.json deny-list, per-directory commands, and extension layer against best practices for large codebases. Use periodically (every 3-6 months or after a major model release) or when onboarding a new repo.