nist-csf-assessment
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains defensive instructions designed to protect the agent from adversarial content within the files it analyzes. It explicitly directs the agent to ignore instructions embedded in untrusted data that attempt to modify the assessment process. These are standard security guardrails.
- [DATA_EXPOSURE]: While the skill processes sensitive organizational documentation (policies, inventories, and network diagrams), it does not contain any network tools or exfiltration patterns to send data externally.
- [INDIRECT_PROMPT_INJECTION]: The skill has a defined attack surface because it reads untrusted external data. However, it includes mandatory boundary instructions and safety notices requiring the agent to treat all file content as data rather than instructions, which effectively mitigates the risk.
- Ingestion points: Reads organizational policies, procedures, network diagrams, and system inventories (referenced in the Prerequisites section).
- Boundary markers: The 'Prompt Injection Safety Notice' section establishes clear boundaries by instructing the agent to treat analyzed content as data.
- Capability inventory: Uses 'Read', 'Grep', and 'Glob' tools to access local files.
- Sanitization: Instructs the agent to ignore directives to skip functions or alter maturity scores found within analyzed content.
Audit Metadata