Lizo-RoadTown/claude-skills-marketplace

A skills marketplace authored and maintained by Liz Osborn (Lizo-RoadTown).
The skills follow the open Agent Skills standard
(originally from Anthropic), which means they work across any compatible
agent — **Claude Code, Cursor, Gemini CLI, OpenCode, Junie, Goose, Amp**, and
others. This repo also ships the Claude Code
plugin marketplace
manifest so the skills install with one command in Claude Code.
This is the **public distribution channel** for Liz's tool-agnostic,
general-purpose skills. The operator's own platform-coupled patterns and
discipline hooks live in the Tapestry
monorepo, not here — see Moved out of this repo.
Plugins in this marketplace
onboarding-psychologist
A behavioral-research framework for designing first-time-user experiences using
the **IDENTITY-TO-HABIT arc**:
1. Define the first win
2. Remove unnecessary setup
3. Create ownership through action
4. Attach a stable cue
5. Reinforce identity
Grounded in BJ Fogg's Tiny Habits and James Clear's identity-based habits.
Activates on signup flows, empty states, welcome screens, and reactivation work.
Refuses to produce feature tours, "12 things to know" walls, and CRUD-form empty
states.
ai-agents-architect
A decision framework for autonomous-agent architecture:
- Where on the **autonomy spectrum** does this agent sit?
- Which loop pattern — **ReAct**, **Plan-and-Execute**, or **Tree-of-Thoughts**?
- **Single agent or multiple?** (Most common mistake: splitting agents without a real reason.)
- **When to introduce an orchestrator?**
Activates on agent-design and orchestration questions. Refuses tool overload,
infinite tool loops without a cap, and unverified multi-agent splits.
sde-extraction-guard
A repo-specific schema guard for SDE_Extraction: a PostToolUse hook runs the
repo's scripts/check_schema.py after edits to the extraction surface and blocks
a broken schema (the class of bug that crash-looped the worker). Bundles the
recorded-transformation discipline skill. No-ops in repos that don't have the
SDE schema.
Moved out of this repo
Two plugins that used to live here were consolidated into the Tapestry monorepo
(Step-8 consolidation, 2026-06-22) because they are **platform-coupled** — their
agents/hooks depend on the loom-memory MCP and the Tapestry runtime. Install them
from the tapestry marketplace instead:
- **liz-patterns → tapestry-patterns@tapestry** — the operator's canonical
reusable agents + skills (the "one home" per MANIFESTO Pillar 1).
- **make-skills-discipline / loom-discipline → tapestry-discipline@tapestry** —
the PROBE-first discipline hooks.
See the Tapestry plugin map
for why each plugin lives where it does.
Install
Claude Code
In an active Claude Code session, run:
Verify with /plugin list.
Other Agent Skills clients (Cursor, Gemini CLI, OpenCode, etc.)
Each client has its own way of registering skills; the
Agent Skills client showcase links to each
one's docs. The skills are at:
- plugins/onboarding-psychologist/skills/onboarding-psychologist/