skill-portfolio-existence-review
An existence/strategy audit for a portfolio of published Claude Code skills, plugin bundles, or adjacent repos — answering the one question the other portfolio skills never ask: should each of these exist at all?
It renders a per-artifact KEEP / CONSOLIDATE / KILL / PIVOT verdict — each with a concrete Recommended Fix — backed by an existence-pinned adversarial panel, web-verified differentiation against competitor/external tools ("is this a skin on a solved problem?"), and redundancy / maintenance-cost / coherence analysis.
🙏 Built on agent-review-panel. This skill is a specialization of that engine for the existence axis — see Built on agent-review-panel below.
The question it answers (and the ones it doesn't)
Most "audit my skills" tooling grades quality, placement, or usage. None asks whether a thing should exist. A flawlessly-built skill that duplicates a free incumbent still shouldn't exist — and this is the skill that will say so.
| If the question is… | Use | Not this skill because… |
|---|
| Should this exist at all? Is it redundant / a reskin / worth my time? | skill-portfolio-existence-review | — |
| Is each skill well-built? (READMEs, badges, factual errors) — and fix them | skill-portfolio-audit | that's quality, not existence |
| Which repo should a skill live in? | skill-portfolio-repo-placement-scan | that's placement, not existence |
| Which skills are unused / dormant? | ecosystem-audit | that's usage, not existence |
| How structurally good is one skill? | schliff | single-skill quality, not portfolio strategy |
| Build or publish a new skill | skill-creator | creation, not audit |
Quick Start
You: i've got 8 skills published on github and i'm not sure half of them should still exist.
help me figure out which to kill or merge.
Claude: [enumerates the repos, reconciles published vs local state, web-verifies the
competitor tools each one overlaps, runs a 5-reviewer existence panel +
fact-check + judge]
## Portfolio Existence Review — should these exist at all?
Existence-coherence: 6.5/10
| Artifact | Verdict | One line |
|------------|-------------|-----------------------------------------------------|
| tool-a | PIVOT | cost-dashboard half is table-stakes vs <incumbent>… |
| tool-b | CONSOLIDATE | byte-identical in two marketplaces; retire one |
| bundle-c | PIVOT | name no longer matches contents; split the track |
| daily-d | KEEP | settled by daily use; only the coupling is live |
You: is my new token-cost dashboard just a reskin of ccusage, or actually different?
You: should i keep building this skill or is it pointless because <tool> already does it?
You: are memory-hygiene and ecosystem-audit redundant? should one absorb the other?
All trigger the same existence review — even without the word "existence".
Installation
Claude Code (plugin install — recommended):
claude plugin marketplace add wan-huiyan/skill-portfolio-existence-review
claude plugin install skill-portfolio-existence-review@wan-huiyan-skill-portfolio-existence-review
Claude Code (git clone):
git clone https://github.com/wan-huiyan/skill-portfolio-existence-review.git ~/.claude/skills/skill-portfolio-existence-review
What you get
The verdict vocabulary — decisive, not hedged:
| Verdict | Meaning |
|---|
| KEEP | Earns its existence as-is. |
| CONSOLIDATE | Merge into / absorb another artifact, or collapse a duplicate publication surface (e.g. retire a second marketplace, keep the repo). |
| KILL | Should not exist; a better tool or a one-liner makes it unnecessary. |
| PIVOT | Has a reason to exist but is aimed wrong; reframe / refocus / rename / re-scope. |
