digitalocean
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill implements operations for managing cloud resources, including the capability to create, delete, and perform actions like rebooting droplets and modifying DNS records through the DigitalOcean API. This represents the core functionality of the skill.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided arguments which are processed and sent to the cloud provider's API, potentially introducing untrusted content into the agent's context from API responses. Ingestion points: src/skill.ts — arguments for tools such as 'droplet-create', 'droplet-list', and 'dns-create'. Boundary markers: None are explicitly defined in the tool outputs to isolate API data. Capability inventory: The skill uses an authenticated HTTP client to perform GET, POST, and DELETE operations against DigitalOcean's API endpoints. Sanitization: Query parameters are escaped using encodeURIComponent, and request bodies are transmitted as structured JSON objects.
Audit Metadata