LLM & Agents

Expert prompt engineering for LLMs and AI agents. Use when users request help writing, reviewing, or improving prompts for chatbots, AI agents, system prompts, instruction sets, or any LLM-based application. Applies research-backed techniques to minimize hallucinations and maximize reliability.

This skill should be used when the user asks to "use delphi", "ask delphi", or wants multiple parallel oracle investigations of the same question to discover divergent insights. Delphi launches multiple oracle agents simultaneously with identical prompts, allowing them to independently explore and potentially discover different paths, clues, and solutions. Results are saved to .oracle/[topic]/ and synthesized into a final document.

Production-ready scripts for Android app testing, building, and automation. Provides semantic UI navigation, build automation, accessibility testing, and emulator lifecycle management. Optimized for AI agents with minimal token output. Android equivalent of ios-simulator-skill.

Central CLAUDE.md file with project guidelines auto-injected into system prompt, achieving 10% accuracy lift in multi-hop reasoning through consistent context. Use for repo-wide context, coding standards enforcement, project guidelines, or architecture documentation. Auto-loaded by Claude Code CLI. Triggers on "project context", "coding standards", "CLAUDE.md", "project guidelines", "repo documentation".

Use when working with any configuration file outside of project directories in ~/Code (unless already managed like ~/Code/CLAUDE.md) - handles chezmoi prefix mappings, template awareness, conflict resolution workflow, and sync to avoid losing changes and breaking templates

Enables AI agents to use Braintrust for LLM evaluation, logging, and observability.Includes scripts for querying logs with SQL, running evals, and logging data.

Update all documentation in .plans, CLAUDE.md files, docs, and .tasks to match current codebase state. Use when user asks to update docs, refresh documentation, sync docs with code, or validate documentation accuracy.

Verwandelt den Agenten in einen professionellen MV3-Architekten und Entwickler mit Fokus auf AI-Integration, Sicherheit, Performance, Testing und Publishing-Compliance.

Execute AI-powered command-line tasks with memory, replay, and resume capabilities using memex-cli. Use when (1) Running AI backend tasks (codex, claude, gemini), (2) Replaying recorded events from JSONL files, (3) Resuming interrupted runs by run_id, (4) Executing prompts with streaming output in jsonl or text format.

Scans codebases and dependencies for known vulnerabilities and security issues

Use multiple Claude agents to investigate and fix independent problems concurrently

Interactive slash command creator for Claude Code. Triggers when user mentions creating commands, slash commands, command templates, command arguments, or wants to build a new command workflow.

Use when executing implementation plans with independent tasks in the current session - dispatches fresh subagent for each task with code review between tasks, enabling fast iteration with quality gates

Comprehensive UX research assistant that analyzes user feedback from surveys, reviews, and interviews to identify trends, cluster themes, surface friction points, and generate actionable product roadmaps. Use when analyzing user feedback, conducting UX research, evaluating product performance, prioritizing features, or transforming qualitative/quantitative feedback into strategic insights.

Use this skill when Claude needs an outside perspective on complex problems, is stuck on a technical challenge, needs validation of an approach, or wants consultation on architectural decisions. Provides access to Codex and Gemini for getting unbiased second opinions and fresh perspectives.

Guidelines for writing unit and integration tests, including mocking strategies and coverage goals.

Analyzes business requirements and translates them into technical specifications

Interactive meal planning wizard. Use when user wants to create a meal plan, optimize their diet, set up nutritional constraints, or find foods that meet their goals. Triggers on requests like "create a meal plan", "optimize my diet", "help me plan meals", "I want to lose weight", "high protein diet".

Singapore corporate secretary and account management guidance. Covers ACRA compliance, statutory filings, AGM/EGM, director duties, and financial record-keeping. Use for compliance calendars, resolution templates, and filing preparation. NOT a replacement for registered filing agent.

Create debug test plan from user input. Use when planning browser debug sessions with Chrome DevTools. Parses prompt for URL/target, identifies interactions to test (clicks, inputs, flows), confirms plan with user via HITL, creates specific test steps (navigate, click, fill, verify). Outputs URL and test plan for debug execution.