vibeteam-readiness
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill accesses sensitive environment variables (e.g., AZURE_API_KEY, GITHUB_TOKEN, SLACK_BOT_TOKEN) and credential files (e.g., .secrets/gmail-credentials.json). These are used to authenticate against the respective services (Azure, Slack, GitHub) to verify system readiness. No data is sent to unauthorized third-party domains.
- [COMMAND_EXECUTION]: Utilizes system tools including curl, kubectl, gh (GitHub CLI), and jq to perform diagnostic checks and process API responses.
- [PROMPT_INJECTION]: The skill processes external data from GitHub issues, Sentry error reports, and Langfuse traces. This content represents an indirect prompt injection surface where maliciously crafted issue titles or error logs could attempt to influence the agent's evaluation process.
- Ingestion points: SKILL.md (via output from curl and gh calls)
- Boundary markers: Absent
- Capability inventory: kubectl, gh, curl, and local file system access
- Sanitization: Absent (uses jq for field extraction without subsequent validation)
Audit Metadata