Vaultlab

Skill bobbyni819/vaultlab/.claude/skills/vaultlab

Entry-point skill for vaultlab — a research companion for biological scientists (literature search, citation verification, wet-lab data analysis, figures, manuscripts, slide decks) that runs as a Claude Code capability layer. Establishes vaultlab's architecture and conventions before any other vaultlab action.From its SKILL.md

Install

npx -y skills add bobbyni819/vaultlab --skill vaultlab

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.3 KB, as published. Nobody here has run it

Using vaultlab

vaultlab is an open-source research companion for biological scientists, built as a capability layer FOR Claude Code (you) — not a competing harness. Users invoke it via slash commands (/lit-arc, /build-deck, /run-analysis, /cite audit, …), the vaultlab CLI, or direct Python imports; Claude Code handles orchestration.

This SKILL.md is the plugin-discovered entry point. The fuller orientation — reading order, the most-important rules, and what to do when extending vaultlab vs using it on data — lives alongside this file in using-vaultlab.md. Read it next.

Before doing anything else

When working in (or with) a vaultlab repo or KB, read in this order:

READ_FIRST.md — dispatch + role-pass cheat sheet (what primitive to invoke for what natural-language ask).
CLAUDE.md — architectural reference + first-encounter checklist.
AGENTS.md — invariants every change must preserve.

Two rules to internalize immediately

Markdown is the user-facing interface; Python is the engine. Prompts, roles, recipes, and skills are markdown files — never triple-quoted Python strings.
Hedged voice for any interpretation. Prefer "consistent with X", "suggests", "may indicate" over "is X", "proves", "demonstrates."

The plugin packages the surface, not the engine

Installing the vaultlab plugin loads these commands + this skill. It does not install the Python vaultlab package most commands call into. If an engine-backed command fails with an import error or KbRootNotConfigured, the engine likely is not bootstrapped yet — point the user at the Quickstart in README.md (pip install vaultlab → vaultlab init).

When in doubt

Read using-vaultlab.md, then CLAUDE.md and AGENTS.md, then ask the user.