domain-dns-ops

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE] (MEDIUM): The skill directs the agent to source ~/.profile and local project profiles to retrieve sensitive credentials such as CLOUDFLARE_API_TOKEN. Accessing shell profiles is a high-privilege action that can expose various environment secrets. The severity is downgraded from HIGH because this access is required for the skill's primary administrative function.
  • [COMMAND_EXECUTION] (LOW): The skill executes local scripts (e.g., bin/namecheap-set-ns) and system tools like cli4, git, dig, and curl. This is the intended behavior but grants the agent significant capability to alter network infrastructure.
  • [PROMPT_INJECTION] (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) as it relies on local data files for its operational logic.
  • Ingestion points: ~/Projects/manager/DOMAINS.md, ~/Projects/manager/DNS.md, and ~/Projects/manager/redirect-worker-mapping.md.
  • Boundary markers: None; the agent is instructed to treat these files as a source of truth without explicit delimiters or warnings to ignore instructions embedded within the data.
  • Capability inventory: cli4 (API access), git (repository management), and arbitrary execution of local binaries.
  • Sanitization: No evidence of content validation or sanitization before the data is processed by the agent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:56 PM