agent_fabric
Verified Safeby nschwermann
Overview
Enables AI agents to safely interact with paid APIs and on-chain workflows on Cronos EVM using scoped, programmable permissions via the Model Context Protocol (MCP).
Installation
pnpm --filter mcp-server devEnvironment Variables
- DATABASE_URL
- SERVER_PRIVATE_KEY
- NEXT_APP_URL
- MCP_CLIENT_SECRET
Security Notes
The project demonstrates robust security practices with the core smart contract design, implementing ERC-7702 delegation with session key support for granular control (least privilege, time bounds, revocability). Sensitive data like session key private keys and API headers are handled using standard, strong hybrid encryption (RSA-OAEP and AES-256-GCM), protecting them at rest and in transit. The API proxy URL validation (blocking localhost/private IPs) helps prevent Server-Side Request Forgery (SSRF) attacks. A potential area for concern lies in the workflow engine's expression resolution (`resolveAllExpressions`). While it does not use `eval` and employs regex for path parsing, if malicious users could create or inject arbitrary workflow definitions, they might theoretically craft expressions to disclose sensitive information from the `WorkflowContext` (which includes wallet addresses and session IDs). However, this risk is mitigated by the fact that workflow definitions are typically created and managed by authenticated users with appropriate permissions. The permissive CORS configuration (`origin: true`) means clients must be diligent in verifying origins, but this is a common setup for public-facing APIs.
Similar Servers
trigger.dev
A platform for building and executing reliable, scalable background tasks and complex workflows, supporting various runtimes (Node.js, Python, Bun), including advanced AI agent orchestration, event-driven processing, and real-time data handling.
evm-mcp-server
Enables AI agents to interact with 60+ EVM-compatible blockchains, perform transactions, query data, and interact with smart contracts via a unified Model Context Protocol (MCP) interface.
AgentUp
A developer-first framework for building, deploying, and managing AI agents, bringing Docker-like consistency and operational ease to AI agent development.
amorce
Provides a secure, cryptographic trust layer and orchestrator for AI agent communication, enabling cross-framework interaction, LLM discovery, and Human-in-the-Loop (HITL) approvals.