saas-finops-optimization
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The 'FinOps Audit Workflow' section directs the agent to check
.envfiles to identify third-party integrations for cost mapping. While necessary for the skill's purpose,.envfiles typically contain highly sensitive API keys and production secrets. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from project manifests and service dashboards, which serves as a potential vector for indirect injection.
- Ingestion points: Local files such as
package.jsonand.env, as well as usage statistics fetched from external provider dashboards (e.g., OpenAI, Vercel). - Boundary markers: The instructions lack specific delimiters or guardrails to isolate the content of these files from the agent's command context.
- Capability inventory: The skill leverages an
ExecutionProxy Interface(using thertkprefix) to perform terminal actions and system tasks. - Sanitization: The skill includes a proactive security instruction to redact all secrets and PII before sharing context with subagents, which acts as a significant mitigation against data exfiltration.
Audit Metadata