devops-engineer

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFECREDENTIALS_UNSAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The file references/kubernetes.md contains a hardcoded connection string with dummy credentials (postgres://user:pass@host:5432/db) in a Kubernetes Secret manifest example. While these are placeholders, they trigger credential exposure patterns.
  • [DATA_EXFILTRATION]: The collect-evidence.sh script in references/incident-response.md automates the collection of highly sensitive information, including network traffic captures via tcpdump, database process monitoring through pg_stat_activity, and full exports of Kubernetes resource configurations. This capability could be leveraged to harvest sensitive environment data.
  • [COMMAND_EXECUTION]: The skill provides numerous scripts and examples that utilize high-privilege CLI tools, including kubectl, terraform, docker, and the GitHub CLI (gh), to manage infrastructure and deployment environments.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8). 1. Ingestion points: User-provided application requirements and environment descriptions in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: High-privilege access via kubectl, terraform, and shell scripts across all files. 4. Sanitization: Absent; user data is interpolated into code templates without validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 08:52 PM
Security Audit — agent-trust-hub — devops-engineer