convez.ai
Skill
kevincui1034

cost-analysis

kevincui1034/cost-analysis

Ver no GitHub
InicialPluginScore · de 0 a 100

Cost & Margin Analysis — an installable Claude Code skill/plugin that audits a project's unit economics: inventories paid surfaces, models pricing, and computes per-plan / per-user margins with redlines flagged.

Instalação

$npx convez-ai kevincui1034/cost-analysis

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 4 meses
Versão
—
Issues abertas
0
Contribuidores
1

README

A tradução para o português deste README ainda está na fila.

Cost & Margin Analysis

A Claude Code skill that audits a project's unit economics. It inventories every paid surface (API calls, SDKs, scrapers, render providers, managed services), attaches per-unit rates from canonical sources, models the pricing structure, and produces per-plan / per-user / per-feature margin math with redlines flagged.

It's stack-agnostic — it detects the project shape first, then runs only the analyses that apply. If the repo has prior cost snapshots in auto-memory, it diffs against them instead of starting cold.

What it does

  • Inventories paid surfaces — Anthropic, OpenAI, Gemini, Replicate, Fal, Apify, Stripe, Resend, Vercel, Supabase, Upstash, Sentry, PostHog, and more.
  • Sources per-unit rates from in-repo rate tables → memory snapshots → (optionally) provider pricing pages via WebFetch, asking before it fetches.
  • Models the pricing structure — subscriptions, top-up packs, usage-based, free tiers, trial/promo credits.
  • Computes three margin scenarios per plan — average user, redlining user (max COGS at min payment), and heaviest legitimate power user — as revenue − COGS = gross margin %, including Stripe fees.
  • Compares across plans and packs and flags non-monotonic or cannibalizing pricing.
  • Diffs against prior snapshots and surfaces margin drift as a finding.
  • Optionally persists a dated memory snapshot and/or a checked-in COSTS.md (both opt-in, never automatic).

Scope note: this skill models recurring, variable run-cost and margins. It does not compute one-time project setup/build costs (dev time, initial provisioning). Fixed monthly infra can be fed into the optional break-even calculator as an input.

Install

This repo is both a Claude Code plugin and a single-plugin marketplace.

Option A — Plugin (recommended)

In Claude Code, run:

/plugin marketplace add kevincui1034/cost-analysis
/plugin install cost-analysis@cost-analysis

Update later with /plugin marketplace update cost-analysis. Because the plugin pins version: 1.0.0, you'll receive changes when that version is bumped.

Option B — Drop the skill in manually

If you'd rather not use the plugin system, copy just the skill folder into your skills directory:

Personal (all projects):

# macOS / Linux
git clone https://github.com/kevincui1034/cost-analysis /tmp/cost-analysis
cp -r /tmp/cost-analysis/skills/cost-analysis ~/.claude/skills/cost-analysis
# Windows (PowerShell)
git clone https://github.com/kevincui1034/cost-analysis $env:TEMP\cost-analysis
Copy-Item -Recurse "$env:TEMP\cost-analysis\skills\cost-analysis" "$env:USERPROFILE\.claude\skills\cost-analysis"

Project-scoped (checked into a repo, shared with collaborators): copy the same skills/cost-analysis folder into your project's .claude/skills/.

Usage

Once installed, invoke it explicitly:

/cost-analysis

…or just ask in natural language — Claude triggers it on prompts like "how much does this cost to run?", "what's our margin?", "are we losing money on the free plan?", "compare the top-up packs", or "is this priced right?".

Repo layout

cost-analysis/
├── .claude-plugin/
│   ├── plugin.json         # plugin manifest
│   └── marketplace.json    # marketplace catalog (one plugin, source "./")
├── skills/
│   └── cost-analysis/
│       └── SKILL.md         # the skill itself
├── LICENSE
└── README.md

License

MIT © Kevin Cui

Ver o original no GitHub

Bom para

Relacionados