The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute shell commands for installation (npm install -g @getdex/cli), setup (mkdir, chmod, echo), and authentication (curl). It also maps MCP tools to a local dex CLI binary for managing CRM data.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and execution of external code by installing the official Dex CLI from npm and configuring an MCP server via npx -y add-mcp https://mcp.getdex.com/mcp -y. These resources are hosted on the vendor's official infrastructure.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection where malicious instructions could be embedded in CRM data.
Ingestion points: Contact descriptions, interaction notes, and custom field values are retrieved from the external Dex database (e.g., via dex_get_contact or dex_list_notes) and presented to the agent.
Boundary markers: The skill does not define specific delimiters to isolate untrusted data retrieved from the API from the agent's instructions.
Capability inventory: The agent has the ability to execute shell commands, perform network operations via curl, and access local files (~/.dex/api-key).
Sanitization: There is no evidence of content sanitization or instruction-filtering for data retrieved from the contact records before it is processed by the agent.

dex-skill