AI & ML

majiayu000 Bundle

Shape, refine, and stabilize human intent into a canonical prompt artifact. Iteratively clarifies ambiguity until user confirms readiness.

majiayu000 Bundle

Apply Chain-of-Verification (CoVe) prompting to improve response accuracy through self-verification. Use when complex questions require fact-checking, technical accuracy, or multi-step reasoning.

majiayu000 Bundle

GPU monitoring and performance metrics for Ollama inference. Check GPU status, VRAM usage, loaded models, and inference performance metrics like tokens per second.

majiayu000 Bundle

Use when building Retrieval-Augmented Generation systems - covers document ingestion, hybrid search retrieval, reranking results, and prompt augmentation for accurate LLM responses grounded in your knowledge base

majiayu000 Bundle

Captures session learnings, decisions, and context to a markdown file for future agent ramp-up. Use when user says "ppage", "page context", "save context", "capture learnings", or before ending a substantial work session.

majiayu000 Bundle

Benchmark one session (or a small recent set) against the rolling average using Agent Monitor data — cost, total tokens, tool count, and workflow complexity score — and report where each metric lands as a percentile of the population. Tells you whether a session was normal, cheap, or an outlier. Use when judging whether a session was typical or out of band.

majiayu000 Bundle

Audit the configured MCP servers (user + project scope) via the Agent Monitor Config Explorer API: transport (stdio vs http), command/args and env variable names, headers, and the source file each definition came from. Reads /api/cc-config/mcp. Use when reviewing MCP integrations for hygiene, duplication, or unexpected transports.

majiayu000 Bundle

Use this skill when the user wants to enable, disable, turn on or off, opt out of, opt in to, or check the status of model-apps telemetry / anonymous usage data. Triggers: "disable telemetry", "turn off telemetry", "opt out of telemetry", "stop collecting usage data", "enable telemetry", "telemetry status".

majiayu000 Bundle

This skill should be used when working with GitHub CLI (gh) for pull requests, issues, releases, and GitHub automation. Use when users mention gh commands, GitHub workflows, PR operations, issue management, or GitHub API access. Essential for understanding gh's mental model, command structure, and integration with git workflows.

majiayu000 Bundle

Verify if a specific action is possible and diagnose any blockers.

majiayu000 Bundle

Run repo startup: prompt load, checklist sync, optional PR triage, worktree hygiene, and CI pull. Best-effort with structured status.

majiayu000 Bundle

Consulta ao corpus RAG do projeto. Busca conhecimento em decisoes, documentacao, learnings e padroes armazenados. Use quando: buscar decisoes passadas, encontrar documentacao, recuperar padroes.

majiayu000 Bundle

Generate descriptive chat titles from the first message using a fast LLM. Runs as a background workflow step after the main response to avoid delaying the experience.

majiayu000 Bundle

Lab Results agent for healthcare workflows.

majiayu000 Bundle

Set up automated agent-driven development with Ralph. Run AI agents in a loop to implement features from user stories, verify acceptance criteria, and log progress for the next agent.

majiayu000 Bundle

Agent Skill，你记一下，我做如下部署调整！

majiayu000 Bundle

Create validated goal-design packets. Triggers: "goal prompt", "goal-design packet", "turn this goal into loop-ready work".

majiayu000 Bundle

Use this when working with BlueSky - fetching threads, reading posts, creating content. Shows you how to use pdsx MCP tools for the task.

majiayu000 Bundle

Mine past agent sessions for working prompts, decisions, and patterns. Use when "what did I ask?", "find that prompt", session archaeology, or agent history.

majiayu000 Bundle

Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling.

majiayu000 Bundle

Define reusable capabilities for an agent, enabling complex workflows and enhancing functionality without code.

majiayu000 Bundle

Production-grade AI agent patterns with MCP integration, agentic RAG, handoff orchestration, multi-layer guardrails, and observability (modern best practices)

majiayu000 Bundle

Test case management for Synnovator platform. Two modes: (1) Guard mode — before any synnovator skill or implementation change, verify existing test cases in specs/testcases/ are not broken. Use when modifying synnovator skill code, endpoint scripts, engine logic, or data model specs. (2) Insert mode — add new test cases to specs/testcases/. Use when user requests adding business scenarios, new feature test coverage, or regression tests. Triggers: 'run tests', 'check testcases', 'add test case', 'verify tests', 'test coverage', or any synnovator implementation change.

majiayu000 Bundle

Use when deploying ANY machine learning model on-device, converting models to CoreML, compressing models, or implementing speech-to-text. Covers CoreML conversion, MLTensor, model compression (quantization/palettization/pruning), stateful models, KV-cache, multi-function models, async prediction, SpeechAnalyzer, SpeechTranscriber.

majiayu000 Bundle

Find Claude Code sessions tracked by the Agent Monitor by project (cwd), model, status, or date, then rank the matches by cost or recency. Pulls the session list and the distinct cwd / facet values so filters use real values rather than guesses. Use when locating a session — "find my EstateWise sessions", "which Opus runs errored this week", "most expensive sessions in /repo".

majiayu000 Bundle

Implementation Agent: Full TDD workflow in a single context window. Writes tests, implements code, and iterates until all tests pass.

majiayu000 Bundle

LibreOJ题目抓取与解题工作流。用户输入形如loj1/loj001或https://loj.ac/p/1 等题号或URL时触发；创建以题号命名的空文件夹，使用Chrome DevTools MCP读取题面并保存为md文件，最后用C++、C、Python、Java、Kotlin、Pascal、Rust、Go生成解答文件。

majiayu000 Bundle

Expert simulation for code guidance and style. Triggers on "code like [expert]", "what would [expert] say", "idiomatic", "best practice", "panel", or domain-specific keywords. Channels documented expert perspectives with explicit confidence.

majiayu000 Bundle

Run with user's arguments:

majiayu000 Bundle

Perform high-recall codebase retrieval using semantic search and symbol indexing. Use when you need to find specific code, understand project structure, or verify architectural patterns before editing.

majiayu000 Bundle

Complete LLM development and engineering skill. Covers strategy selection (prompting vs fine-tuning vs RAG), dataset design, PEFT/LoRA fine-tuning, evaluation workflows, vLLM deployment, and production optimization. Modern best practices for building, evaluating, and scaling LLM systems.

majiayu000 Bundle

Inspect or change Babysitter model-routing policy by phase.

majiayu000 Bundle

Generate a commit message for changes already in the Git staging area, and await approval from the user. Use when the agent needs to generate a commit message for staged changes. Only generates the commit message, but does not perform the commit.

majiayu000 Bundle

RAG-first ORCHESTRATOR startup. Minimal context (~500 tokens), queries RAG on-demand. Token-efficient alternative to startupO.

majiayu000 Bundle

Smart router — reads a task description and routes to the right agent or skill. Use when the user runs /studio, describes a task without naming a specific skill, or asks which skill to use.

majiayu000 Bundle

Privileged applier that LANDS meta-optimize / corpus-audit patches the user approved — the ONLY skill permitted to mutate the skill corpus from a self-modification proposal, with cross-model jury and human approval at landing. Use when the user says "meta apply", "/meta-apply", "land the staged patches", "应用优化", after a /meta-optimize run.