vap-mcp-server
Execution control layer for AI agents - Reserve, execute, burn/refund pattern for media generation
README
VAP – Execution Control Layer for AI Agents
"VAP is where nondeterminism stops."
The Problem
If your agents call paid APIs directly, you don't have:
- Cost control – No budget limits, no spending caps
- Retry limits – Failed calls can loop indefinitely
- Failure ownership – No clear accountability when things go wrong
Your AI agent needs to generate an image. It calls DALL-E. The call fails. It retries. Fails again. Retries 10 more times.
You just burned $5 on nothing.
The Solution
VAP is an Execution Control Layer that sits between AI agents and paid external APIs.
It enforces:
- Pre-commit pricing – Know exact cost before execution
- Hard budget guarantees – Reserve → Burn → Refund model
- Deterministic retry behavior – No runaway costs
- Explicit execution ownership – Every task has an owner
How It Works
Agent: "Generate an image of a sunset"
↓
VAP: "That will cost $0.18. Reserving..."
VAP: "Reserved. Executing..."
VAP: "Success. Burning $0.18. Here's your image."
If it fails:
Agent: "Generate an image of a sunset"
↓
VAP: "That will cost $0.18. Reserving..."
VAP: "Reserved. Executing..."
VAP: "Failed. Refunding $0.18. Error: Provider timeout"
Your agent never sees the complexity. It just gets deterministic results.
Pricing
| Type | Preset | Price |
|---|---|---|
| Image | image.basic |
$0.18 |
| Video | video.basic |
$1.96 |
| Music | music.basic |
$0.68 |
| Campaign | streaming_campaign |
$5.90 |
| Full Production | full_production |
$7.90 |
No surprises. No variable pricing. No "it depends."
MCP Integration
VAP is on the official MCP Registry: io.github.elestirelbilinc-sketch/vap-e
Claude Desktop Configuration
{
"mcpServers": {
"vap": {
"url": "https://api.vapagent.com/mcp",
"transport": "streamable-http"
}
}
}
Available Tools (10)
| Tool | Description |
|---|---|
generate_image |
Generate AI image ($0.18) |
generate_video |
Generate AI video ($1.96) |
generate_music |
Generate AI music ($0.68) |
estimate_cost |
Get image generation cost |
estimate_video_cost |
Get video generation cost |
estimate_music_cost |
Get music generation cost |
check_balance |
Check account balance |
get_task |
Get task status by ID |
list_tasks |
List recent tasks |
execute_preset |
Execute named preset |
SDK Usage
Installation
pip install vape-client
Basic Usage
from vape_client import VAPClient
client = VAPClient(api_key="your_api_key")
# Cost is pre-committed: $0.18
result = client.generate_image(
prompt="A serene mountain landscape at sunset"
)
print(f"Image URL: {result.url}")
print(f"Cost: ${result.cost}")
Async Usage
import asyncio
from vape_client import AsyncVAPClient
async def main():
client = AsyncVAPClient(api_key="your_api_key")
# Budget enforced, retries limited
result = await client.generate_image(
prompt="A futuristic cityscape"
)
print(f"Image URL: {result.url}")
asyncio.run(main())
API Endpoints
| Endpoint | Method | Description |
|---|---|---|
/v3/generate |
POST | Create media execution task |
/v3/tasks/{id} |
GET | Retrieve task status |
/v3/tasks/{id}/result |
GET | Retrieve task result |
/v3/balance |
GET | Check account balance |
Full API Docs: api.vapagent.com/docs
The Three Guarantees
1. Pre-Commit Pricing
Every task has a known cost before execution. No surprises.
2. Budget Enforcement
Set a max budget. VAP enforces it. Hit the limit? Task rejected. Balance protected.
3. Failure Ownership
Every task has an explicit owner. Every failure has an address. No more "the agent did something and I don't know what."
Links
- MCP Registry: registry.modelcontextprotocol.io
- API Documentation: api.vapagent.com/docs
- MCP Endpoint:
https://api.vapagent.com/mcp
License
MIT License – see the LICENSE file for details.
VAP – Execution Control Layer for AI Agents
"VAP is where nondeterminism stops."
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
E2B
Using MCP to run code via e2b.