Sold by: Floating Cloud
Deployed on AWS
Free Trial
Agent Provost: Guardrails for AI trading. An enterprise-grade Model Context Protocol (MCP) proxy for the Alpaca API. Features a hot-reloadable trading rules engine, strict rate limiting, and immutable WORM audit logging to S3 for complete compliance.
Overview
Connecting AI agents (like Claude or ChatGPT) to live trading APIs introduces massive security and compliance risks. Agent Provost solves this by acting as a secure, zero-trust Model Context Protocol (MCP) proxy specifically designed for the Alpaca trading API. It sits directly between your AI and your brokerage account, ensuring that LLMs can only execute trades within strict, pre-defined boundaries.
Built for fintech developers and algorithmic traders, Agent Provost features a hot-reloadable Rules Engine that enforces maximum trade notionals, share limits, and explicit symbol allowlists or blocklists. It includes robust rate limiting to prevent runaway AI loops and API bans.
Security is baked into the architecture. The appliance is 100% stateless, pulling ephemeral credentials into a secure RAM disk on every boot. For compliance and auditing, all AI trading requests and system actions are immutably logged to an AWS S3 bucket with Object Lock (WORM) enabled, ensuring a tamper-proof audit trail of every AI decision.
Enterprise Pricing & Private Offers: The public pricing reflects our standard on-demand rate. However, we strongly encourage enterprise customers to contact us for a Private Offer. Customers willing to commit to a multi-year agreement or participate in a logo-sharing case study are eligible for up to a 50% discount on annual licenses. Please contact support@floatingcloud.io to negotiate a Private Offer before deploying.
Highlights
- Customizable Trading Rules Engine: Enforce strict guardrails on AI agents with hot-reloadable policies. Set maximum trade notionals, share limits, and explicit symbol allowlists or blocklists to prevent rogue AI trading behavior.
- Zero-Trust Ephemeral Architecture: Designed for maximum security, the appliance is 100% stateless. API keys are never written to disk; they are fetched dynamically from AWS Secrets Manager into a secure RAM disk on every boot.
- Immutable WORM Audit Logging: Achieve enterprise-grade compliance with automatic, tamper-proof logging. Every AI request, proxy action, and system event is streamed directly to an Object Lock-enabled Amazon S3 bucket.
Details
Sold by
Categories
Delivery method
Delivery option
Agent Provost Secure Deployment
Latest version
Operating system
Ubuntu 24.04
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Free trial
Try this product free for 30 days according to the free trial terms set by the vendor. Usage-based pricing is in effect for usage beyond the free trial terms. Your free trial gets automatically converted to a paid subscription when the trial ends, but may be canceled any time before that.
Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time. Alternatively, you can pay upfront for a contract, which typically covers your anticipated usage for the contract duration. Any usage beyond contract will incur additional usage-based costs.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimension | Cost/hour |
|---|---|
m6g.medium | $5.00 |
t4g.micro | $5.00 |
t4g.medium | $5.00 |
t4g.small | $5.00 |
t4g.large | $5.00 |
m6g.large | $5.00 |
Dimensions summary
You pay by the hour based on the EC2 instance size you run. All six options bill per hour and change only in compute capacity. The t4g family (t4g.micro, t4g.small, t4g.medium, t4g.large) offers ARM-based instances that scale from lighter to heavier workloads. The m6g family (m6g.medium, m6g.large) provides more balanced compute and memory. You choose the instance that fits your expected trading traffic. Hourly billing means your cost tracks the running time of the instance you select. No upfront commitment applies under this usage model.
Top-of-mind questions for buyers
Each unit is one running EC2 instance-hour for the size you pick. The proxy runs on that instance inside your own AWS account. You pay for each hour the instance stays running. The t4g and m6g sizes differ only in compute and memory capacity.
Software charges meter running time only. A stopped instance does not accrue hourly software fees. Because the appliance is stateless and holds no credentials on disk, you can stop it safely. Note that underlying AWS storage or other resources may still bill separately from this software charge.
No. The hourly charge covers only the running instance. Audit trails stream to an Object Lock-enabled Amazon S3 bucket in your own account, and that S3 storage bills through AWS separately. The proxy deploys via a CloudFormation Template inside your cloud perimeter.
floatingcloud.io
Helpful?
Vendor refund policy
All sales are final. We do not offer refunds or prorated refunds for any hourly or annual subscriptions. The Software is provided "AS IS" and Licensee assumes all risks associated with its use. You may cancel your subscription at any time to prevent future billing, but no refunds will be issued for past usage or prepaid annual terms.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Agent Provost Secure Deployment
This CloudFormation template provisions the complete, zero-trust architecture required for Agent Provost. It creates a secure AWS Secrets Manager entry to hold your Alpaca API keys and trading rules, ensuring they are never written to the EC2 hard drive. It also provisions an Amazon S3 bucket with Object Lock (WORM) enabled to store your immutable trading logs. Finally, it launches the Agent Provost EC2 instance with a least-privilege IAM role, allowing it to securely fetch its configuration into a temporary RAM disk on boot.
CloudFormation Template (CFT)
AWS CloudFormation templates are JSON or YAML-formatted text files that simplify provisioning and management on AWS. The templates describe the service or application architecture you want to deploy, and AWS CloudFormation uses those templates to provision and configure the required services (such as Amazon EC2 instances or Amazon RDS DB instances). The deployed application and associated resources are called a "stack."
Version release notes
Update of Agent Provost. Includes the core MCP proxy, very strict hot-reloadable Rules Engine, and immutable WORM logging to S3.
Run the cloudformation, answer the questions with deep thought. after this Fill out the ALL of the deployment parameters including the rules: Alpaca Credentials: Enter your Alpaca API Key and Secret Key (stored securely in AWS Secrets Manager, never on disk). Provost Token: Create a secure, random password. Your AI will use this to authenticate with the proxy. Trading Rules: Set your MaxTradeNotional, MaxSharesPerTrade, and your Symbol Allowlists/Blocklists. Wait for the stack status to reach CREATE_COMPLETE. Go to the Outputs tab of your CloudFormation stack to find the Public IP Address of your new appliance. Step 2: Connect Your AI (MCP Client Setup) Agent Provost acts as a remote MCP server. Update your MCP client configuration file to route traffic to your EC2 instance's IP address, using the PROVOST_TOKEN you created during deployment.
For Claude Desktop Edit your claude_desktop_config.json file:
{ "mcpServers": { "alpaca-provost": { "type": "sse", "url": "http://<YOUR_EC2_PUBLIC_IP>:8000/sse", "env": { "PROVOST_TOKEN": "<YOUR_PROVOST_TOKEN>", "PROVOST_USER": "claude-desktop", "PROVOST_MACHINE": "work-laptop" } } } } For Cursor Add this to your Cursor MCP settings (.cursor/mcp.json):
{ "mcpServers": { "alpaca-provost": { "type": "sse", "url": "http://<YOUR_EC2_PUBLIC_IP>:8000/sse", "env": { "PROVOST_TOKEN": "<YOUR_PROVOST_TOKEN>", "PROVOST_USER": "cursor-ide", "PROVOST_MACHINE": "dev-machine" } } } } (Note: Replace <YOUR_EC2_PUBLIC_IP> and <YOUR_PROVOST_TOKEN> with your actual values. The PROVOST_USER and PROVOST_MACHINE headers are optional but highly recommended, as they will be recorded in your immutable S3 audit logs to identify exactly who initiated the trade).
Step 3: Verify the Connection Restart your MCP client (Claude or Cursor). Open a new chat and type: "What is my current account balance and buying power?" Test the Rules Engine: Ask the AI to buy 10,000 shares of a stock. Agent Provost will intercept the request, block it based on your CloudFormation rules, and log the blocked attempt to your S3 bucket.
Additional details
Usage instructions
- Launch the CloudFormation stack and fill in your Alpaca API keys and trading rules.
- Wait for the stack to reach CREATE_COMPLETE.
- Connect to the EC2 instance using AWS Systems Manager (Session Manager).
- The proxy is automatically running on port 8000.
- To update your trading rules, simply update the CloudFormation stack parameters. The instance will automatically sync the new rules within 10 minutes without requiring a reboot.
Resources
Vendor resources
Support
Vendor support
Support is provided via our open-source GitHub repository and email. For bug reports, feature requests, and community support, please open an issue at https://github.com/CharmingSteve/agent-provost . For direct deployment assistance or critical inquiries, contact us at support@floatingcloud.io . We aim to respond to all inquiries within 72 hours.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
The Qualys Cloud Agent provides continuous security assessment and compliance monitoring
Free, open-source starter pack to deploy a production-ready AI agent in minutes, not weeks. Pre-configured with Claude Haiku 4.5, web search, and code execution. Perfect for CEOs or technical teams building agentic AI solutions.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.