vault
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill enables the agent to process and analyze screenshots of Vault audit logs, login errors, and UI banners to diagnose authentication issues.
- Ingestion points: The agent is instructed to read audit/login error screenshots and sealed UI banners (referenced in
gemini.md). - Boundary markers: There are no explicit delimiters or specific instructions for the agent to treat text extracted from images via OCR as untrusted data.
- Capability inventory: The skill enables the agent to generate HCL policies (
SKILL.md,gpt.md), provide infrastructure-as-code snippets, and suggest bash commands for secret management. - Sanitization: The skill includes a mandatory directive to redact visible tokens found in screenshots from its summaries, which addresses data exposure but does not explicitly mitigate malicious instructions embedded in the image text.
- [SAFE]: The skill references official documentation and resources from HashiCorp's developer portal and GitHub repositories for technical guidance.
- [SAFE]: Example bash scripts and HCL policies provided in the skill use standard parameterization and placeholders, following secure coding practices.
Audit Metadata