matthull
UserCurated Claude Code skills with plugin architecture
Categories
Indexed Skills (41)
code-review
Structured code review with parallel specialist agents in separate subagent contexts. Uses /skill-discovery for domain-aware reviews. Each leg runs in its own fresh subagent for clean analysis.
direct-handoff
Hand off the current task to a fresh Claude session in a terminal-multiplexer sidecar pane. This skill should be used when context is getting heavy (>50%), when a task needs a fresh context window, or when the user explicitly asks to hand off work. Handles doc updates, ephemeral context capture, task list transition, and sidecar launch with confirmation.
handoff
Create handoff documents for tightly scoped implementation tasks sized to fit in a single Claude Code context window (target 50-60% max context). Handoffs embed core mandates, verification loops, and spec context so they can be passed to dedicated short-lived subagents that implement the task and exit. Designed for subagent dispatch (pass handoff as prompt to a Task tool subagent).
pr-create
Create or update a pull request with a rich, structured description. Supports two modes: (1) orchestrated — receives a pre-filled PR description from a verifier or other source, (2) standalone — gathers context from git and fills the template itself. Handles PR mechanics: create, update, push, reviewer assignment, ticket linking.
qa-plan
QA strategy and classification. Works in two modes — PR-scoped (analyze a diff) or feature-scoped (analyze a feature from docs, specs, and code exploration). Classifies verification items by execution channel, detects gaps, and produces a structured QA plan or test matrix. Use standalone or as the verification phase of a development pipeline.
rca
Root cause analysis for technical investigations. This skill should be used when diagnosing production bugs, service degradation, unexpected behavior, or any issue requiring systematic evidence gathering. Triggers on "investigate", "root cause", "why is this failing", "diagnose", or when an issue requires gathering data from multiple sources before drawing conclusions. Emphasizes data-driven reasoning, context gathering, and not jumping to conclusions.
react-native-best-practices
Provides React Native performance optimization guidelines for FPS, TTI, bundle size, memory leaks, re-renders, and animations. Applies to tasks involving Hermes optimization, JS thread blocking, bridge overhead, FlashList, native modules, or debugging jank and frame drops.
skill-creator
Create, adapt, and audit Claude Code skills that conform to the plugin architecture. This skill should be used when authoring a new skill, migrating an existing skill to be portable across environments, or reviewing a skill for architectural conformance. Covers the frontmatter convention, capability declarations, semantic verbs, the resolution template, extension points, and graceful degradation.
explore
AI-guided interactive code exploration for holistic comprehension. Pre-analyzes a codebase area with parallel subagents, then guides an interactive session with focused diagrams, code links, and narrative. Use when you want to deeply understand how a chunk of a system works — not audit it, not review it, but comprehend it. Triggers on "explore", "walk me through", "explain this codebase", "how does this work", or when the goal is understanding existing code.
nvim-copilot
Neovim copiloting protocol — drive a side-by-side neovim instance as a presentation surface for code. Manages setup, navigation, annotations, floating windows, code snippet popups, tour maps, and communication routing between Claude Code and neovim panes. Use as a dependency from any skill that shows code alongside conversation (exploration, code review, debugging, teaching). Also invocable directly to start a copiloted neovim session.
orchestrate
Plan and execute multi-step work through composable modes and sequences. Triages operator goals into concrete completion criteria and agent/operator boundaries, selects or composes orchestration patterns, and manages delegation across agents and teams. Teaches operators orchestration principles by modeling them in practice. Use when work needs structured coordination — not for simple single-agent tasks.
research
Deep research with parallel subagents, automatic citations, and chain-of-evidence footnotes. This skill should be used when conducting research that requires verifiable sources, when creating documents that need footnote citations, or when working with downstream documents that must maintain traceability to primary sources. Triggers on "research", "investigate", "find sources for", or when creating documents that require citations.
supabase-postgres-best-practices
Postgres performance optimization and best practices from Supabase. Use this skill when writing, reviewing, or optimizing Postgres queries, schema designs, or database configurations.
nt-comms
Review workplace communications against common workplace norms before sending. Use for sensitive work messages, emails, or chat - especially when navigating accountability, feedback, or politically charged situations. Runs as a subagent so the review is not already adapted to the writer's own style.
threads
Conversational scaffolding for multi-threaded exploratory discussions. This skill should be used during shaping sessions, brainstorming, spec conversations, or any exploration where multiple threads emerge. Tracks questions, ideas, decisions, agreements, and topics so nothing is lost. Dependency for PM and spec skills.
meal-planning
Meal planning, recipe management, and shopping list creation. Covers the full arc from "what sounds good?" through recipe extraction, ADHD-friendly formatting, shopping lists, and pinned Keep notes for watch access. Triggers on "meal plan", "what should I eat", "recipe", "shopping list", "grocery", or when the artifex wants help with food for the week.
resume-workflow
Generate and manage resumes in JSON Resume format using the resumed CLI. Use when updating resume content, rendering to HTML or PDF, or trying different themes.
bash-scripts
Bash scripting safety patterns, bats-core TDD testing, Docker execution, idempotency. Load this skill when writing or modifying shell scripts.
davinci-resolve
DaVinci Resolve 20 video editing for developer screencasts. This skill should be used when editing OBS recordings into railscasts-style screencasts — importing footage, removing/replacing audio, speed ramping, voiceover, captions, and export. Triggers on "davinci", "resolve", "screencast", "edit video", "edit recording", or video post-production tasks.
disk-cleanup
Analyze disk usage and clean package and container caches on an Arch Linux workstation. Use when the user mentions disk space, storage, cleanup, or that a drive is full.
fastmcp-python
FastMCP/Python MCP server development — tool design, naming conventions, pytest, tmux-sidecar E2E testing. Load this skill when building or modifying MCP servers.
go-cli
Go CLI development patterns — testing strategy, argument parsing, error handling, concurrency, and quality gates. Load this skill when implementing or modifying Go CLI tools.
markdown-to-pdf
Convert markdown files to beautifully formatted PDFs with dark mode support. This skill should be used when users want to export markdown documents to PDF format, especially when they need dark themes, proper ASCII diagram rendering, or professional formatting for tables with icons.
ruby-rails
Ruby/Rails patterns — fixture builder, RSpec organization, migrations, console verification. Load this skill for Ruby on Rails development tasks. Use during coding phases like IMPLEMENT as well as technical architecture phases like DESIGN.
spec-check
Spec compliance checker. Given a spec and a branch, compares every requirement against the implementation diff with evidence. Reports per-requirement PASS/FAIL with file:line references and an overall verdict. Use standalone after any spec-driven implementation, or as the verification phase of an orchestrated pipeline.
supabase-migrations
Guidance for writing Supabase database migrations with idempotent patterns. This skill should be used when creating or modifying database objects (tables, functions, RLS policies, triggers, indexes) or any SQL committed as a migration file.
testing
Test scenarios, coverage analysis, test quality standards, and the mandatory test failure stop protocol. Load this skill for any task requiring test coverage.
waydroid-adb
Control Waydroid Android emulator via ADB for Expo/React Native development. This skill should be used when needing to take screenshots of Waydroid, verify app rendering, install APKs, or interact with the Android emulator. Triggers on tasks involving Android emulator verification, mobile app visual testing, or Waydroid troubleshooting.
calcinatio
Refinement through verifying force — the principle and practice of subjecting work to fires that burn away what doesn't hold up. Covers deriving fires from witnesses, context shaping, and the orchestration patterns (dialectical, manifold). Load at session start; calcinatio applies to all work that will reach a witness.
citations
Standard citation format for human- and LLM-friendly source traceability across all document types. Covers web sources, internal sources (Slack, meetings, voice notes), and chain-of-evidence propagation. This skill defines the format convention — not invoked directly, but referenced by other skills that produce cited documents.
manual-qa
Claude QA Engineer procedures — automated AND "manual" testing via tools before human UAT. Load this skill when Claude must verify features end-to-end via tools rather than deferring to human.
orthogonal-emanation
Divergence before convergence — generate maximally independent candidates to escape mode collapse. Use when choosing among approaches and premature convergence would be costly (UX design, architecture, research angles, communication strategies).
product-intent-signal-extraction
Extract product intent signals from raw inputs (meeting transcripts, Slack threads, voice notes, brain dumps) into a structured signal map with per-source quality assessment. Designed to run as a subagent so extraction doesn't blow up the orchestrator's context. This skill should be used when pm-spec or any product management skill needs to process raw signals before shaping — never invoked directly by users.
project-management-signal-extraction
Extract project management signals from raw inputs (meeting transcripts, Slack threads, stand-up notes, async updates) into a structured signal map. Designed to run as a subagent so extraction doesn't blow up the orchestrator's context. This skill should be used when a project management skill or orchestrator needs to process raw signals before synthesis — never invoked directly by users.
request-research
Create research request documents for topics requiring investigation. Produces portable documents usable by any research agent. Use when a feature or decision needs external research before proceeding.
skill-discovery
Discover and load relevant skills for the current task. Scans the installed skill catalog, evaluates relevance, and outputs reasoning. Use at the start of any orchestrated agent work to ensure the right practice skills are loaded.
slack-thread-triage
Scan a Slack channel, classify threads by action type (product, project, technical, no-action), and tag them with emoji reactions. Already-tagged threads are skipped on subsequent runs. This skill should be used when triaging Slack channels for actionable threads, or as part of a recurring triage workflow.
spec
Create or update a feature specification from a natural language feature description. Enforces brownfield-first discovery, traceability tags, research coordination, and incremental document creation. Use when starting design work on a new feature or enhancement.
tmux-sidecar
Create and interact with secondary tmux panes or windows for bidirectional command execution. This skill should be used when "tmux sidecar" is mentioned, or when needing to run interactive sessions (REPLs, rails console, database clients, log tailing) in a separate pane or window while maintaining control from the main session.
supabase-sql
Supabase SQL application logic — RLS policies, auth hooks, triggers, pgTAP testing. Load this skill for Supabase database work.
youtube-transcribe
Transcribe a YouTube video and do something with the transcript. Wraps youtube-to-signal. Use when a YouTube URL appears and the user wants to work with its content — summarize, analyze, extract quotes, review, or anything else.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.