AI & ML

shawn-sandy Bundle

Optimizes what a project loads into every coding-agent session. Counts which skills, plugins, slash commands, and subagents past sessions actually invoked, prunes the rest from the load set, then tightens CLAUDE.md against the same data by deleting dead references, replacing ignored always-do rules with hooks, and pinning the skills that carry the dev cycle in a short table. Use when the user asks to optimize their agent setup or project context, prune unused skills or plugins, audit skill usage, or slim down CLAUDE.md.

cgbarlow Bundle

Retrieve and faithfully reproduce DoView Planning handbook diagrams to accompany an outcomes-theory answer. Use after a doview-outcomes-answer response when the user asks to see the diagrams, requests the visual model, says "show me the picture/diagram/figure", or whenever an outcomes-theory answer would land better with the original tool diagrams visible. Reproduces Mermaid blocks from the local Markdown edition first (https://github.com/cgbarlow/doview-book), and falls back to the upstream PNG/image-file URLs from doviewplanning.org when a chapter's diagram does not translate cleanly to Mermaid. Faithful — never redraws, simplifies, paraphrases, or substitutes images or Mermaid blocks. Pair with doview-outcomes-answer, which sets up the seed list this skill consumes.

cgbarlow Bundle

Answer outcomes-theory questions strictly from Dr Paul Duignan's DoView Planning and Outcomes Theory Handbook (2025). Use whenever the user asks for an outcomes-theory perspective, requests a DoView analysis, asks "what does outcomes theory say about…", asks to analyse a strategy/proposal/plan/policy/programme/business-case/M&E plan/contract through an outcomes lens, mentions outcomes theory or outcomes systems or DoView Boards or strategy-outcomes diagrams or theory-of-change, shares a document and wants it critiqued for clarity of outcomes, or wants a formal standalone response with the Summary + Full structure and raw-visible copy-safe URLs that the handbook's Prompt A v1.1.9 demands. Strictly avoids first-person and drafting-advice language; uses outcomes theory as the primary frame and DoView as its applied form. Pair with `doview-image-retriever` to retrieve the original diagrams (Mermaid from this repo or PNG from doviewplanning.org).

jmmarotta Bundle

Guide for creating effective skills. Use when creating, updating, restructuring, or validating a skill package with specialized workflows, knowledge, tool integrations, or bundled resources. Use `agent-instruction-writing` when editing the instruction text itself.

jmmarotta Skill

Spawn subagents with the bonsai coding-agent CLI when no task tool or native subagent is available. Use for self-contained exploration, research, review, parallel investigations, or isolated work that would bloat the parent context.

jmmarotta Skill

Use when editing `AGENTS.md`, `SKILL.md`, or agent system prompts.

nikolai-vysotskyi Skill

Use trace-mcp tools for code navigation, impact analysis, and framework-aware queries instead of Read/Grep/Glob/Bash. Activate whenever the agent needs to explore, understand, or modify a codebase that has trace-mcp indexed.

nikolai-vysotskyi Skill

Use trace-mcp apply_codemod for any bulk mechanical change instead of repeated Edit calls. Activate whenever the same edit pattern would be applied 2+ times, across one file or many.

nikolai-vysotskyi Skill

Safe refactoring workflow using trace-mcp — assess risk, find candidates, check impact, and rename symbols across all files without missing import sites or cross-file references.

microsoft Bundle

Convert Prompt Flow flow definitions to Microsoft Agent Framework (MAF) workflows. Parses flow.dag.yaml, maps nodes to Executors, and generates runnable Python code using agent-framework 1.0.x. WHEN: convert promptflow, migrate promptflow, promptflow to MAF, promptflow to agent framework, convert flow.dag.yaml, migrate flow to MAF, convert PF flow, PF to agent-framework, convert DAG flow to workflow, migrate LLM flow. DO NOT USE FOR: writing new MAF workflows from scratch (no source flow), deploying MAF workflows (use maf-online-endpoint), enabling tracing (use maf-tracing), or general agent-framework Q&A.

ukgovernmentbeis Skill

Land a PR that requires a coordinated ts-mono submodule change. Use when the check-schema-and-types CI job fails, when a Python model change requires regenerating the OpenAPI spec / TypeScript types, or when any change touches the ts-mono submodule and needs to be landed.

gentleman-programming Skill

PR creation workflow for Agent Teams Lite following the issue-first enforcement system. Trigger: When creating a pull request, opening a PR, or preparing changes for review.

gentleman-programming Skill

Creates new AI agent skills following the Agent Skills spec. Trigger: When user asks to create a new skill, add agent instructions, or document patterns for AI.

gentleman-programming Skill

Issue creation workflow for Agent Teams Lite following the issue-first enforcement system. Trigger: When creating a GitHub issue, reporting a bug, or requesting a feature.

modelscope Bundle

Access ms-agent's advanced AI capabilities via MCP tools: deep research, document research, financial research, code generation, video generation, web search (arxiv/exa/serpapi), LSP code validation (TypeScript/Python/Java), concurrent-safe file editing, and agent delegation. All project-level capabilities support async submit/check/get patterns. Use when the user asks to research a topic, analyze documents, generate code or videos, validate code, edit files, or delegate tasks. Requires ms-agent (pip install ms-agent).

hwanyong Skill

Control the VS Code Debugger (DAP) via CLI commands using mcp-debug-tools. Use when debugging, setting breakpoints, stepping through code, or inspecting variables.

ukgovernmentbeis Bundle

Retrieve detailed Linux system information including OS distribution, kernel version, CPU model and core count, memory usage, and uptime. Use when the user asks about system specs, hardware details, RAM, processor info, kernel version, or needs a host inventory summary.

nvidia Bundle

Troubleshoot cuOpt LP/MILP problems including errors, wrong results, infeasible solutions, performance issues, and status codes. Use when the user says something isn't working, gets unexpected results, or needs help diagnosing issues.

nikgor Skill

Guided implementation workflow for SFUMATO modules. Use when starting to implement any module (service, tool, endpoint, model, Jinja2 template), when unsure about dependency order, or when you want to verify architecture compliance before writing code. Encodes project-specific patterns and common mistakes to avoid.

nikgor Skill

Research prompt engineering best practices for AI image generation and LLM-as-judge scoring patterns. Use when writing or improving Jinja2 prompt templates in app/prompts/, when implementing the optimize loop, when judge scores cluster too narrowly, or when prompt revision fails to improve results across iterations.

kodustech Skill

End-to-end front→back performance debugging for the Kodus web app. Use when a screen is slow, blank, looping, or you need to trace a UI perf problem down to the API query. Drives Playwright/Chrome MCP to open + measure the screen, then reads the API use-case/repository and the DB (Postgres/Mongo indexes, explain) to find and fix the root cause, verifying live.

kodustech Skill

Use when the user wants Kodus to review local changes, run `kodus review` or `--prompt-only`, fix Kodus review findings, or check commit, push, or merge readiness.

kodustech Bundle

Use when the user wants Kodus to review local changes through a local CLI build or non-production API such as `node dist/index.js`, localhost, or QA, including local `--prompt-only` flows.

miles990-skillpkg Bundle

Help users configure MCP (Model Context Protocol) servers for Claude Code and other AI platforms.

alexander-danilenko Bundle

Write or audit technical text against ASD-STE100 Simplified Technical English so a tired, non-native reader understands it on the first pass: 20-word procedural and 25-word descriptive sentence limits, imperative steps, condition before command, one word for one meaning, and only the approved modals can/will/must. Use for READMEs, runbooks, procedures, error messages, release notes, incident reports, API guides, UI copy, and agent instructions. Trigger on: "STE", "Simplified Technical English", "ASD-STE100", "controlled language", "plain English", "make this clearer", "write for non-native readers", "prep this for translation", "cut the fluff", or any request to check text for compliance. Prefer this over general copy-editing whenever the text instructs, documents, or warns.

toruai Skill

Look up how Claude Code itself works — hooks, skills, subagents, MCP, plugins, settings — from the official docs, so answers come from the current spec instead of memory.

toruai Skill

Branch and commit conventions that keep an agent's work reviewable — one branch per task, main stays clean, nothing force-pushed.

yida29 Skill

Llm Backend

youdotcom-oss Skill

Additional code review guidelines specific to this codebase. Auto-loaded by the review agent during PR reviews to enforce project conventions.

youdotcom-oss Skill

Review and improve AGENTS.md or scoped agent instructions in this repo. Use when asked to compress, clarify, reorganize, or harden agent workflow rules.

luoandorder Skill

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

luoandorder Skill

Use when the user wants to build an implementation plan and a matching `/goal` prompt to drive Claude Code through it autonomously. Triggers on phrases like "plan and goal", "/plan-and-goal", "build a plan and a goal prompt", or when the user asks to scope a task into a plan file plus a goal directive. Produces a plan markdown file in /tmp and a self-contained goal prompt that points back at the plan.

luoandorder Skill

Use when the user wants to brainstorm, iterate on a draft, plan an approach, or think something through together. Triggers on phrases like "let's think through", "help me plan", "brainstorm with me", "let's iterate", or when the user shares a draft/idea and asks for input. Drives the conversation through opinionated AskUserQuestion calls anchored in personas. Optional user-supplied steering input is treated as an extra prompt that shapes personas and focus.

nvidia Skill

Use when fixing NeMo Agent Toolkit documentation path-check failures, especially failed `ci/scripts/path_checks.py` output, slash-delimited text mistaken for paths, relative path references, Markdown code escaping, and path-check allowlist decisions.

nvidia Skill

Use before creating, editing, or deciding whether to update any AI coding agent skill in this repository, including corrections to existing skill behavior, references, or routing.

nvidia Bundle

Use when configuring or running NeMo Agent Toolkit optimization with `nat optimize`, including Optuna parameter tuning, prompt evolution, optimizer sizing, output interpretation, and optimizer datasets.