Agent-Skills.md

Agent Skills: V1 Menu

Use when explicitly choosing which v1tamins skill fits a task. Triggers on "which v1 skill", "v1 menu", or "/v1-menu".

UncategorizedID: v1-io/v1tamins/v1-menu

Repository

v1-io/v1tamins
v1-ioLicense: MIT
61

Install this agent skill to your local

pnpm dlx add-skill https://github.com/v1-io/v1tamins/tree/HEAD/plugins/v1tamins/skills/v1-menu

Skill Files

Browse the full folder contents for v1-menu.

Download Skill

Loading file tree…

plugins/v1tamins/skills/v1-menu/SKILL.md

Skill Metadata

Name
v1-menu
Description
Use when explicitly choosing which v1tamins skill fits a task. Triggers on "which v1 skill", "v1 menu", or "/v1-menu".

V1 Menu

You don't remember every v1 skill, so ask. Read the user's situation, then point them at the right skill or chain below — recommend one, name the runner-up, and say why. This menu is an index, not documentation: each skill's own SKILL.md holds the detail. If the question is what to work on next rather than which skill fits, use /v1-what-next.

Skills marked (explicit) never fire autonomously — the user must name them. They are the easiest ones to forget, so surface them whenever they fit.

Building something new (idea → shipped)

  1. /v1-interview-me — office-hours questioning to flesh out an idea before planning. Hands off to /v1-learning-from-customers when the question is real demand, or /v1-diagnosing-constraints when a stuck process is the real problem.
  2. /v1-prd — turn the settled idea or ticket into requirements. /v1-bare-bones — strip an overscoped plan to the smallest useful version. /v1-strategy-review — challenge assumptions and ambition before committing.
  3. Build: /v1-implement-unit (explicit) takes one adequately planned ticket through isolated implementation, fresh review-board cycles, and a mergeable PR. For manual work, drive /v1-write-tests for new coverage and /v1-refine before calling it done.
  4. Ship: /v1-pr (explicit) (create a draft PR), /v1-pr-description (title/body), /v1-land-pr (explicit) (commit → push → CI → ready for review; never merge), /v1-prove-work (explicit) (record browser proof for the PR).

Reviewing code

  • /v1-deep-review — review a branch or PR (code, docs, or config) for merge risk and structural maintainability.
  • /v1-pr-walkthrough — create a dated self-contained HTML explanation of a PR with Background, Intuition, Code, and interactive Quiz sections.
  • /v1-review-board (explicit) — propose current peers/models first, wait for a roster, then run a read-only multi-agent ledger.
  • /v1-address-review — resolve review comments that already exist.
  • /v1-phone-a-friend (explicit) — propose one current OAuth-backed counterpart and wait for approval before launching it.

Diagnosing problems

  • /v1-debug — root-cause diagnosis of any observable expected-versus-actual gap (code or not).
  • /v1-diagnosing-constraints — Theory of Constraints when throughput, queues, or WIP are the problem.
  • /v1-designing-habit-systems — design or diagnose a habit, routine, or cadence that fails to stick.

Fixing and improving code

  • Broken: /v1-fix-tests (failing suite), /v1-e2e-testing (browser/E2E and flakes).
  • Working but rough: /v1-refine (quality pass / deslop / hindsight rewrite on the diff or named files).

Research and analysis

  • /v1-deep-research — multi-source research report. /v1-autoresearch-skill (explicit) — autonomous optimize-measure-keep loop against a measurable target.
  • /v1-learning-from-customers — demand evidence and discovery. /v1-testing-prototypes — user tests for prototypes.
  • /v1-reviewing-data-graphics — audit charts and dashboards. /v1-reviewing-usability — review a UI or flow for user error.

Communicating and documenting

  • /v1-skin-suit — draft or lightly edit natural prose for its reader and channel while preserving meaning and requested voice.

  • /v1-stickify — make copy memorable. /v1-html-it — self-contained HTML page or report. /v1-md2docs (explicit) — markdown to Google Doc.

  • /v1-explainer-video — narrated animated explanation of a concept, process, or PR using your available renderer and speech provider.

  • /v1-changelog — what shipped, from merged PRs. /v1-docs-freshness — sync docs after changes.

Working on skills and prompts

  • /v1-skilling-it — manage an Agent Skill from Canonical Source creation, editing, audit, and validation through separately gated packaging, installation, or publication. /v1-canon2skill — mine source material for candidate skill ideas before authoring. /v1-goldpan (explicit) — pan recent PRs and sessions for compound-worthy lessons.
  • /v1-prompt-engineering — work on general prompts, system prompts, hooks, or sub-agent briefs for any model or host, including GPT-5.5 and OpenRouter specifics; use /v1-skilling-it when the artifact is an Agent Skill.
  • /v1-shared-language — extract a domain glossary from the conversation.