managing-digitalocean-app-platform

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes bash scripts to execute doctl and jq commands for resource discovery and analysis.
  • Evidence: Bash code blocks in the 'Phase 1: Discovery' and 'Phase 2: Analysis' sections of SKILL.md.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and processing external data from DigitalOcean without sanitization or boundary markers.
  • Ingestion points: Command outputs from doctl apps list, doctl apps get, and doctl apps logs in SKILL.md.
  • Boundary markers: None present in the prompt instructions.
  • Capability inventory: Shell execution via bash, JSON parsing with jq, and cloud resource management via doctl as defined in SKILL.md.
  • Sanitization: Absent; the skill passes raw command output directly to the agent for analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 06:01 AM
Security Audit — agent-trust-hub — managing-digitalocean-app-platform