fedramp
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [SAFE]: The skill consists exclusively of informational markdown documents providing guidance on FedRAMP compliance and certification.
- [EXTERNAL_DOWNLOADS]: The skill references official government documentation and templates from fedramp.gov and nist.gov. These are well-known, trusted sources for federal compliance information.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions (in SKILL.md) represent a surface for indirect prompt injection as they require the agent to ingest and analyze user-supplied system architectures and compliance descriptions. 1. Ingestion points: User-provided CSO descriptions and architecture details processed during gap assessments. 2. Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are provided. 3. Capability inventory: The skill does not define or use any external tools, scripts, or network capabilities. 4. Sanitization: No sanitization or filtering of user input is specified. Because the skill lacks any executable capabilities, this surface does not present a significant security risk.
- [NO_CODE]: No scripts, executables, or dynamic context injection patterns were found. The skill operates entirely through natural language instructions.
Audit Metadata