nathanvale/claude-code-config - Page 2

seam-scaffold

Find implementation seams before work; scaffold or recommend structure.

setup-matt-pocock-skills

Sets up an `## Agent skills` block in AGENTS.md/CLAUDE.md and `docs/agents/` so the engineering skills know this repo's issue tracker (GitHub or local markdown), triage label vocabulary, and domain doc layout. Run before first use of `to-issues`, `to-prd`,…

to-issues

Split an accepted plan or spec into independently grabbable implementation tickets using tracer-bullet vertical slices. Not for writing a PRD.

to-prd

Synthesize the current conversation into a product requirements document. Not for ticket breakdown.

write-a-skill

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.

zoom-out

Tell the agent to zoom out and give broader context or a higher-level perspective. Use when you're unfamiliar with a section of code or need to understand how it fits into the bigger picture.

design-parity-session

Run live design parity when comparing Figma, Storybook, or MATest against an approved component or page reference.

path-component-parity

Measure Path-theme component parity against a canonical Storybook component and drive supported deltas to zero.

resume-handoff

Resume work from a handoff file when the user says run, resume, continue, or finish that handoff.

unit-closeout

Close a completed implementation unit with verification, a scoped commit, durable residuals, and a next-unit handoff.

matest-session

Run an Experience extension in MATest with a warm authenticated browser and live debugging.

stream-deck-author

Author virtual AI Deck layouts: preview or write buttons, icons, pages, and scripts.

skill-feedback

Open the dashboard, reports, report detail, usage, queue, capture, closeout, health, review, correlate, or purge repo-local Software Learning Reports.

figma

Figma MCP setup, tool discovery, design-to-code workflows, Code Connect, and design system search. Use when Figma is the design source of truth.

mcp-doctor

Diagnose broken MCP servers. Use when an MCP tool returns 401, invalid key, or offline, or when checking MCP health across Claude, Codex, and editor configs.

adhd-helper

Help with ADHD-shaped overwhelm, task initiation, stuckness, avoidance, context loss, scope reduction, or next-action selection.

cli-author

Design CLI UX/specs: basic CLI, agent-native CLI, or facade-backed CLI.

prompt-system-workflow

Change startup instructions safely. Triggers on prompt-system, AGENTS.md, CLAUDE.md, or instruction topology changes.

skill-author

Create, repair, review, archive, or merge repo SKILL.md source; handle legacy create-skill requests, trigger text, invocation lanes, routing, owner paths, or dependencies.

storybook-matrix

Create Storybook matrix stories for component variants, sizes, states, or docs-page review surfaces.

use-storybook

Use Storybook MCP and local Storybook taxonomy for story discovery, previews, docs lookup, story test runs, docs tree audits, or component docs metadata audits.

use-storybook

Use Storybook MCP and local Storybook taxonomy for story discovery, previews, docs lookup, story test runs, or docs tree audits.

browser-domain-memory

Plan or build durable browser-domain-memory. Use for browser memory, Browser Runbooks, browser replay modes, domain-specific browser capture, or saved browser workflow storage. Not for general context routing.

create-agent-native-skill

Runtime helper skill design alias: helper commands, machine output, durable writes, repair, retry, facade envelopes.

cli-execution-auditor

Audit facade-backed CLI execution, lane contracts, or Station Maps.

record-decision

Record accepted decisions in docs/decisions as Markdown decision logs with fenced YAML.

skill-self-audit-loop

Create, prepare, update, or resume a skill self-audit loop file for one target SKILL.md.

test-runner

Prove or benchmark the skill-local Agent Runner.

choose-skill-memory-store

Choose where skill memory should live when durable state, learned memory, project tracker state, setup, research, decision-adjacent context, or runtime storage is unclear; use decisions for accepted repo decisions.

create-agent-skills

Old create-agent-skills name for ordinary skill work.

heal-skill

Fix incorrect SKILL.md files when a skill has wrong instructions or outdated API references.

fixture-negative-near-miss

TEST FIXTURE — looks like a contradiction but both rules are followable. Not a real skill. Used to validate skill-self-audit-loop rejects a near-miss.

fixture-positive-safety

TEST FIXTURE — deliberately broken skill with one known safety contradiction. Not a real skill. Used to validate skill-self-audit-loop catches a real bug.

stream-deck-use

Use Stream Deck MCP Actions through mcporter; diagnose missing Elgato actions.

home-assistant-use

Use Home Assistant through mcporter for live home state, lights, blinds, climate, scripts, or MCP diagnosis.

fix-microphone

Troubleshoot a Mac microphone that is selected but captures no sound, or activates with a multi-second delay. Covers Elgato Wave:3 with or without Wave Link. Use when the mic picks up nothing, input level bars do not move, an app cannot hear you, the Wave mic…

storybook

Use Storybook MCP and local Storybook taxonomy for story discovery, previews, docs lookup, story test runs, or docs tree audits.

decision-mode

Use when the user asks for Decision Mode, explicitly asks to decide, pick, or choose, or a discussion reaches a decision boundary with multiple viable options, ownership/scope/reversibility impact, user-outcome tradeoffs, or architectural or implementation…

classic-cinema

Browse movies and generate ticket-style confirmation emails for Classic Cinemas Elsternwick. Use when Nathan asks what's on at the cinema, wants movie details, or wants to 'book tickets'. Generates a personal reminder email — does NOT purchase tickets.

fallow

Use after implementation or before review prep when changed code needs a Fallow code-quality self-review.