npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@hypershub/cli

v0.1.10

Published

HypersHub local integration configurator CLI

Vulnerabilities

Powered byPowered by Snyk
  • 0High
  • 0Medium
  • 0Low

Links

Git

Maintainers

simonguosimonguo

Keywords

hypershubclicodexclaude-codeopencode

Readme

HypersHub CLI

CI npm version license: MIT

hy imports HypersHub integration settings into local AI coding tools, switches default models, and verifies connectivity.

Repository

Install

npm install -g @hypershub/cli
hy --help

Requirements: Node.js 18+.

Quick start

hy init codex
hy init claude-code
hy init opencode
hy config set baseUrl https://apiclaw.cc
hy config set apiKey hy-xxx
hy models
hy check all
hy check all --live
hy use gpt-5.5
hy use claude-deepseek-v4-pro claude-code --live
hy test codex
hy doctor

Non-interactive examples:

hy init codex --url https://apiclaw.cc --key hy-xxx --model gpt-5.4 --yes
hy init claude-code --url https://apiclaw.cc --key hy-xxx --model claude-sonnet-4-6 --yes
hy init opencode --url https://apiclaw.cc --key hy-xxx --model gpt-5.4 --yes

Global config

Use hy config to save common settings once. This avoids entering the API key every time you run hy models, hy init, or other commands. hy init also writes the confirmed Base URL, API key, and selected model back to this global config, so future commands can reuse them automatically.

hy config set baseUrl https://apiclaw.cc
hy config set apiKey hy-xxx
hy config set defaultModel gpt-5.5
hy config list
hy config get apiKey
hy config path
hy models

Supported keys:

  • baseUrl: HypersHub base URL. You can use https://apiclaw.cc or https://apiclaw.cc/v1; the CLI normalizes it automatically.
  • apiBaseUrl: OpenAI-compatible API URL. Setting this also updates baseUrl.
  • apiKey: HypersHub API key.
  • defaultModel: default model used by commands when --model is omitted.

Precedence, from highest to lowest:

command flags > environment variables > hy config > interactive prompt/defaults

Environment variables:

HYPERSHUB_BASE_URL=https://apiclaw.cc
HYPERSHUB_API_KEY=hy-xxx
HYPERSHUB_DEFAULT_MODEL=gpt-5.5

Config file locations are cross-platform:

  • macOS/Linux default: ~/.hypershub/config.json
  • Linux with XDG_CONFIG_HOME: $XDG_CONFIG_HOME/hypershub/config.json
  • Windows default: %APPDATA%\HypersHub\config.json

Run hy config path to see the exact path. API keys are redacted by default in output. Use --show-secrets only when you explicitly need the full value. The file is written with 0600 permissions where the platform supports it.

Base URL normalization

You can enter either form when prompted or when using --url:

https://apiclaw.cc
https://apiclaw.cc/v1

The recommended input is https://apiclaw.cc. The CLI automatically normalizes URLs for each integration:

  • Codex / OpenCode / /v1/models / /v1/responses: uses https://apiclaw.cc/v1.
  • Claude Code / ANTHROPIC_BASE_URL: uses https://apiclaw.cc; Claude-compatible requests are sent to /v1/messages automatically.

So users do not need to remember which tool needs /v1.

Daily workflow

Most users only need to initialize once, then use hy use for day-to-day model switching:

hy init all                 # one-time setup
hy check all --live         # verify local config and model availability
hy use gpt-5.5              # switch all configured tools
hy use gpt-5.5 codex        # switch one tool only
hy use claude-deepseek-v4-pro claude-code --live

hy use reads the existing local config, reuses the saved Base URL/API Key, and updates only the default model. This is the recommended way to switch models after setup.

Commands

  • hy init codex: writes ~/.codex/config.toml and ~/.codex/model-catalogs/all-models.json.
  • hy init claude-code: updates shell profile and ~/.claude/settings.json with ANTHROPIC_BASE_URL, ANTHROPIC_AUTH_TOKEN, default model variables, and related settings.
  • hy init opencode: writes ~/.config/opencode/opencode.json.
  • hy init all: applies all supported integrations.
  • hy models: queries /v1/models and lists all models available to the API key. It reads hy config automatically. Use --json for scripts.
  • hy config list|get|set|unset|path: manages global CLI settings such as baseUrl, apiKey, and defaultModel.
  • hy check [target|all]: inspects local generated configs, redacts API keys, and reports missing URL/key/model/catalog fields.
  • hy check [target|all] --live: additionally calls /v1/models and verifies the configured default model exists for the current key.
  • hy use <model> [target|all]: switches the default model in existing local configs without re-entering URL/API Key. Use --live to verify availability first. If target is omitted, it applies to all configured integrations.
  • hy test <codex|claude-code|opencode>: sends a small test request to /v1/responses or /v1/messages.
  • hy doctor: checks common config files and environment variables.

Dynamic model catalog

For hy init codex and hy init opencode, the CLI calls:

GET /v1/models
Authorization: Bearer hy-xxx

The generated local model catalog only contains models available to that API key. If the request fails, the CLI falls back to a small built-in catalog so initialization can still complete.

Switch models

After hy init, use hy use to change the default model without re-entering the Base URL or API Key:

hy use gpt-5.5                    # switch all configured integrations
hy use gpt-5.5 codex              # switch Codex only
hy use claude-deepseek-v4-pro claude-code
hy use deepseek-v4-pro opencode
hy use gpt-5.5 all --live         # verify /v1/models before writing

What it updates:

  • Codex: ~/.codex/config.toml, and ensures the model exists in ~/.codex/model-catalogs/all-models.json.
  • Claude Code: shell profile plus ~/.claude/settings.json, including cc-switch-compatible model keys.
  • OpenCode: ~/.config/opencode/opencode.json.

Use --json for scripts:

hy use gpt-5.5 codex --live --json

Restart the target app after switching. For Claude Code, reload your shell if needed:

source ~/.zshrc

Platform notes

macOS / Linux / WSL

hy init claude-code updates ~/.claude/settings.json as well as one of:

  • ~/.zshrc
  • ~/.bashrc
  • ~/.config/fish/config.fish

The ~/.claude/settings.json sync is used for compatibility with Claude Code and cc-switch style model overrides, so /model should show the model selected during hy init. Reload your shell after initialization, for example:

source ~/.zshrc

Windows PowerShell

hy init claude-code writes a HypersHub block to the PowerShell profile and sets user-scoped environment variables when that profile is loaded.

Default profile path:

~/Documents/PowerShell/Microsoft.PowerShell_profile.ps1

After initialization, restart PowerShell or run:

. $PROFILE

Safety

  • Existing files are backed up by default with .bak.YYYYMMDDHHmmss suffix.
  • --no-backup disables backup.
  • --dry-run prints generated config without writing files.
  • API keys are redacted in dry-run output and prompts.
  • Config files are written with 0600 permissions where the platform supports it.
  • .env files are not published.

Query, check, and test

After configuration, use these commands to find problems quickly:

hy config set apiKey hy-xxx
hy models
hy models --json
hy check codex
hy check all --live
hy use gpt-5.5
hy use claude-deepseek-v4-pro claude-code --live
hy test codex --model gpt-5.4
hy test claude-code --model claude-sonnet-4-6

hy check --live is the most useful first diagnostic: it reads the local config, calls /v1/models, and confirms the configured model is actually available to that API key. Use hy use <model> --live when the config is valid but you want to switch models safely.

Troubleshooting

/v1/models fails during init

The CLI will continue with fallback models. Check:

  • URL should normally be https://apiclaw.cc or https://apiclaw.cc/v1.
  • API key should start with hy-.
  • Network/DNS can reach your HypersHub gateway.

Claude Code still uses Anthropic directly

Run:

hy doctor

Then reload the shell profile. The CLI also synchronizes ~/.claude/settings.json; if another switcher changes the model later, run hy use <model> claude-code or re-run hy init claude-code. If you previously logged into Anthropic inside Claude Code, run /logout in Claude Code and restart the terminal.

Codex does not show custom models

Re-run:

hy init codex

Then restart Codex Desktop/CLI. The model catalog is written to:

~/.codex/model-catalogs/all-models.json

Development

git clone [email protected]:hypershub/hypershub-cli.git
cd hypershub-cli
npm test
npm run pack:check
npm run smoke:pack
npm run preflight

See also:

Publish beta:

npm publish --access public --tag beta

Publish dry-run:

npm publish --dry-run --access public

Publish stable via GitHub Trusted Publisher:

  1. Bump package.json version and update CHANGELOG.md.
  2. Commit and push to main.
  3. Create and push a tag, for example v0.1.9.
  4. Publish a GitHub Release for that tag, or run the Publish workflow manually.

The workflow uses npm Trusted Publisher / OIDC, so it does not need an NPM_TOKEN secret. prepublishOnly still runs npm run preflight before publishing.