The Agent Skills Directory

[DATA_EXFILTRATION]: The skill performs legitimate file and network operations required for its social media management functionality.\n
The media:upload command in src/commands/media.ts reads local files and uploads them to the vendor's infrastructure using presigned URLs.\n
Authentication credentials are stored locally in the user's home directory (~/.zernio/config.json) and accessed by the client to sign API requests to zernio.com.\n
Contact data can be read from local JSON files for bulk creation tasks using the contacts:bulk-create command.\n- [INDIRECT_PROMPT_INJECTION]: The skill interacts with social media content, which serves as a potential source of untrusted input for an AI agent.\n
Ingestion points: Commands such as inbox:messages, inbox:comments, and inbox:reviews in src/commands/inbox.ts retrieve external content.\n
Boundary markers: The skill does not define specific delimiters or instructions to the agent to treat this data as untrusted.\n
Capability inventory: The skill allows the agent to publish content, send messages, and create automations across 14 social platforms.\n
Sanitization: The skill returns raw content from the platform APIs without explicit sanitization by the skill code itself.\n- [EXTERNAL_DOWNLOADS]: The skill relies on official vendor packages and well-known libraries.\n
It utilizes the official @zernio/node SDK for API communication.\n
It uses standard Node.js utilities such as open and yargs for CLI functionality.

zernio