Back to search
claude-code-effort
MarketplaceWorks with Claude CodeGitHub
Pick Claude Code models and effort levels from data: where the spend goes on this machine, and which configs are worth their cost per task.
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Checked today
Badge for your README
Markdown
[](/listings/jinhuang712--claude-code-effort)HTML
<a href="/listings/jinhuang712--claude-code-effort"><img src="/api/listings/jinhuang712--claude-code-effort/badge.svg" alt="findsafeskills safety rating" /></a>Install
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
claude plugin marketplace add jinhuang712/claude-code-effort && claude plugin install cc-effort@claude-code-effort
Requirements, update and removal are in [INSTALL.md](INSTALL.md).
---
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
.claude-plugin/marketplace.json · Manifest · 23 lines
- {
- "name": "claude-code-effort",
- "owner": {
- "name": "Jin Huang",
- "url": "https://github.com/jinhuang712"
- },
- "metadata": {
- "description": "Model and effort choices for Claude Code, from usage and benchmark data",
- "version": "0.1.0"
- },
- "plugins": [
- {
- "name": "cc-effort",
- "source": "./",
- "description": "Usage analysis, benchmark frontier and a model/effort review: /cc-effort:review",
- "category": "productivity",
- "tags": [
- "cost",
- "models"
- ]
- }
- ]
- }
README.md · README · 65 lines
- # claude-code-effort
- > Pick Claude Code models and effort levels from your own data.
- A Claude Code plugin with one command, `cc-effort`, and one skill, `/cc-effort:review`. **It measures your own transcripts, not a guess:** where the money went on this machine, which model and effort configs are worth their cost per task, and which standing instructions pick a model or an effort.
- ---
- ## What it does
- - 📊 **Local spend**: cost by model, effort and role, per turn and per sub-agent run, read from Claude Code's transcripts on this machine.
- - 🏁 **Benchmark frontier**: scores and cost per task from [Artificial Analysis](https://artificialanalysis.ai/), ranked on a Pareto frontier where a near-tie goes to the cheaper token.
- - 🔍 **What is in force**: the settings that pick a model and an effort, and every instruction line in CLAUDE.md, memory or agent definitions that names one.
- - 🧭 **A review**: the skill reads all of it, recommends a model and effort per role, and edits settings and CLAUDE.md only after you agree.
- ---
- ## Install
- ```
- claude plugin marketplace add jinhuang712/claude-code-effort && claude plugin install cc-effort@claude-code-effort
- ```
- Requirements, update and removal are in [INSTALL.md](INSTALL.md).
- ---
- ## How to use
- In a Claude Code session, run the skill or ask in your own words:
- > /cc-effort:review
- > Which model and effort should my sub-agents use?
- The session fetches the benchmark table, measures your transcripts, ranks the configs and answers with one config per role. The numbers come from commands you can run yourself in that session:
- ```
- cc-effort frontier --tie 3
- ```
- ```
- config index $ / task min / task verdict
- haiku-5-5 xhigh 34.6 $0.25 9.5 frontier
- haiku-5-5 max 36.5 $0.41 14.8 beaten by haiku-5-5 xhigh
- sonnet-5-5 medium 30.8 $0.78 3.8 beaten by haiku-5-5 high
- sonnet-5-5 high 41.5 $1.64 7.2 frontier
- ```
- ---
- ## Docs
- - [FEATURES.md](FEATURES.md)
- - [CLI.md](CLI.md)
- - [INSTALL.md](INSTALL.md)
- - [CHANGELOG.md](CHANGELOG.md)
- - [specs/METHOD.md](specs/METHOD.md)
- - [skills/review/SKILL.md](skills/review/SKILL.md)
- ---
- ## License
- MIT. See [LICENSE](LICENSE). Benchmark data belongs to [Artificial Analysis](https://artificialanalysis.ai/); the plugin fetches it on your machine and does not ship it.
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.