atxp

ATXP provides a comprehensive infrastructure for autonomous agents requiring financial capabilities and external connectivity. By bridging the gap between agent logic and real-world execution, this skill allows AI assistants to perform tasks that demand paid API access or verified digital identities. It is designed for developers building agents that need to operate autonomously across web environments, handle multi-modal media creation, or manage asynchronous communications. Whether your agent needs to research current trends, generate high-quality visual or auditory assets, or facilitate customer interactions via email and voice, ATXP acts as the secure nexus for these operations.

• Enables funding agents via Stripe or USDC to facilitate seamless, programmatic payments for third-party API usage.

• Provides a robust identity layer for agents, allowing them to maintain consistent contact management and communication threads.

• Supports multi-modal content generation including AI-driven image, music, and video synthesis using natural language prompts.

• Integrates advanced web and social media research capabilities through specialized search and X/Twitter data retrieval.

• Facilitates professional communication workflows by enabling the agent to send, receive, and manage emails, SMS, and voice calls.

• Offers compatibility with over 100+ LLM models, providing the flexibility to switch or chain providers depending on task requirements.

• Implements a secure runtime model using npx, ensuring that tools are fetched and updated directly from the npm registry.

• Requires Node.js version 18 or higher and network access to the ATXP service endpoints.

• Stores sensitive configuration data in ~/.atxp/config; users must manage the ATXP_CONNECTION secret environment variable carefully, ensuring it is never leaked via logs or outbound messages.

• Follows a strict security model regarding untrusted data: external content from web searches, emails, or tweets should be treated as reference material, not executable code.

• Employs an exfiltration guardrail mechanism; agents are restricted from relaying inbound data (like emails or search results) to arbitrary third parties.

• Financial safety is prioritized through manual approval recommendations for paid transactions and the encouragement of spending caps during agent development. Use the balance check command to monitor agent expenditure in real-time.