Total 56,867 skills, AI & Machine Learning has 9450 skills
Showing 12 of 9450 skills
Framework-independent LLM serving benchmark skill for comparing SGLang, vLLM, TensorRT-LLM, or another serving framework. Use when a user wants to find the best deployment command for one model across multiple serving frameworks under the same workload, GPU budget, and latency SLA.
Use a local QMD knowledge base through UXC over MCP stdio, with daemon-backed session reuse and typed retrieval flows that avoid repeated model warmup and unnecessary query-expansion latency.
The root entry of the CodeStable workflow family — introduces the overall system to users and routes users' specific requests to the correct cs-* sub-skills. Trigger scenarios: users only input `cs` / `/cs`, say "introduce codestable", "do something with codestable", "I want to do X, which skill should I use", "don't know which one to use", or users' described requests are open-ended (e.g., "start working") and haven't converged to a specific sub-skill. This skill itself **does not perform actual tasks** — it doesn't write specs, write code, or read/write content products in the codestable/ directory — it only performs scanning, routing, prompting, and then transfers control to the target sub-skill.
Nassim Taleb's Antifragility framework applied to a business idea, system, or portfolio position. Spawns a team of specialist agents — Fat-Tail Detector, Fragility Auditor, Optionality Scout, Iatrogenics Checker, Skin-in-the-Game Auditor — who each apply a distinct lens from Taleb's Incerto to evaluate whether the subject is fragile, robust, or antifragile. The lead synthesizes into a convexity assessment: what's the payoff structure under disorder, where are the hidden tail risks, and the honest Taleb verdict. Use when the user says "taleb this", "is this fragile", "antifragility analysis", "what would Taleb think", "tail risk check", or proposes a business/system and wants structural risk analysis. Works standalone or after /munger for complementary analysis.
3-에이전트(Architect→Builder→Reviewer) 루프로 단일 기능을 설계·구현·검증하는 팀 스킬. "3a로 만들어줘", "3에이전트", "설계-구현-검토", "team-3a" 키워드로 트리거. peach-team보다 가벼운 단일 기능·소규모 수정에 적합.
Run comprehensive agent-native architecture review with scored principles
Expert skill for using Future AGI — the open-source end-to-end platform for evaluating, observing, and improving LLM and AI agent applications with tracing, evals, simulations, datasets, gateway, and guardrails.
Analyze community opinions from forums and comment sections. Scrapes comments from Bilibili, Reddit, or GitHub Issues, clusters them by semantic similarity, and extracts the core arguments, debates, and viewpoints. Produces a structured report showing what the community actually thinks — not just a summary of comments, but the underlying positions people hold and where the real disagreements are. Use this skill when the user wants to understand public opinion on a topic, find the main points of contention in a discussion, or do competitive/event research from community sources. Triggers include requests to "analyze comments", "what are people saying about X", "summarize the debate", "find the key arguments", "what's the community consensus", or any task involving opinion extraction from forum or comment data.
Build command-line interfaces for AI agents. Covers arguments, flags, subcommands, help text, output formats, error messages, exit codes, config/env precedence, and safe/dry-run behavior. Use when building a new CLI or refactoring an existing one for agent use.
Quick summary of the last session — commands run, files changed, and what to do next.
Generate and evaluate solution options
Create validated LLM-as-a-Judge evaluators following best practices — binary Pass/Fail judges with TPR/TNR validation for measuring specific failure modes. Use when you need to automate quality checks, build guardrails, or measure a specific failure mode identified during trace analysis. Do NOT use when failures are fixable with prompt changes (use optimize-prompt) or when failure modes are unknown (use analyze-trace-failures first).