JSPM

  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
  • Downloads 251
  • Score
    100M100P100Q108742F

Package Exports

  • @findagent/cli

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@findagent/cli) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

@findagent/cli

Install AI agents into Claude, ChatGPT, Gemini, Cursor, and VS Code — straight from your terminal.

npm version license

findagent is the command-line installer for the FindAgent cross-LLM agent marketplace. One command writes the exact files an agent needs into your client — recipe agents and tool-using "doer" agents alike — and the result is byte-identical to the guided web installer.

What is FindAgent?

FindAgent is a marketplace for AI agents that work across LLM clients. Agents are declarative: a manifest describes an agent's prompt and tools, and a single trusted runtime executes them — agents never ship code to run on your machine. Browse the registry, install an agent, and it shows up in Claude, ChatGPT, Gemini, Cursor, or VS Code.

Install

# Run on demand — no global install needed
npx @findagent/cli <command>

# …or install it globally
npm i -g @findagent/cli

Requires Node.js ≥ 18.

Quickstart

# Sign this machine in (browser device flow)
findagent login

# Install an agent into Claude (default target)
findagent install pr-reviewer

# Pin a version and target Cursor
findagent install pr-reviewer@1.2.0 --llm cursor

# Preview what would be written — no filesystem changes
findagent install pr-reviewer --llm claude --dry-run

Commands

Command What it does
findagent install <slug[@version]> Install an agent into your LLM runtime. Flags: --llm <target>, --dir <path>, --dry-run.
findagent download <department> Download a Department (a multi-agent team) to run locally. Flags: --dir <path>, --dry-run.
findagent list (aliases: ls, installed) List the agents you have purchased or installed.
findagent login Authenticate this machine via the OAuth device flow. Flag: --token <jwt> to paste a token instead.
findagent logout Remove the stored FindAgent credentials.
findagent whoami Show the currently signed-in account.
findagent secrets set <ref> <value> Store a per-agent external credential (e.g. an API token).
findagent secrets list List stored credential refs (values hidden).
findagent secrets rm <ref> Remove a stored credential.

Run findagent --help for the full flag reference.

--llm targets

Target Aliases Lands in
claude claude-code .claude/agents/
chatgpt openai, gpt .findagent/chatgpt/
gemini google .findagent/gemini/
cursor — .cursor/mcp.json
vscode vs-code, code .vscode/mcp.json
custom generic .findagent/

The target set is the canonical cross-LLM list shared with the marketplace and @findagent/schema (LLM_TARGETS), so the picker, the install bundle, and the CLI never drift. A recipe agent lands as a per-client file; a hosted/MCP agent lands as a ready mcpServers connector config.

Credentials

If an agent calls an external service, its manifest declares credential slots. Store a value for a slot locally — it never leaves your machine, and the runtime only attaches it to requests whose host the slot allows:

findagent secrets set <ref> <value>
findagent secrets list
findagent secrets rm <ref>

Secrets are stored owner-only (mode 0600) in your OS config dir, shared with the @findagent/mcp runtime.

Configuration

Env var Default Purpose
FINDAGENT_API https://beta.findagent.cloud API base (override for staging).
NO_COLOR — Disable ANSI color output.

Credentials live in the OS config dir:

  • macOS / Linux: $XDG_CONFIG_HOME/findagent/auth.json (or ~/.config/findagent/)
  • Windows: %APPDATA%\findagent\auth.json

How it works

The CLI and the web installer at /install/<slug> produce the identical install artifact. Both call the same marketplace endpoint (GET /api/v1/agents/<slug>/install-bundle?llm=<target>), whose response shape is the shared installBundleResponseSchema in @findagent/schema. The server renders the manifest and per-LLM files; the CLI decodes the returned files[] and writes them verbatim.

Paid agents are gated exactly like the web flow: an unpurchased paid agent returns 402 and the CLI refuses with a clear message rather than writing anything.

findagent login (without --token) runs an OAuth 2.0 Device Authorization Grant (RFC 8628): it prints a short user_code, opens the /device confirmation page, and completes the poll automatically once you approve in the browser.

License

MIT