accounting-software-selection
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized network operations were detected. The skill implements strong internal guardrails, such as prohibiting the invention of data and requiring human verification for vendor claims.
- [INDIRECT_PROMPT_INJECTION]: The skill inherently possesses an ingestion surface for untrusted data, as it processes user-supplied software requirements and evaluation scores into generated artifacts.
- Ingestion points: User responses regarding software shortlists, business requirements, and pricing (documented in Step 1 and Step 2 of
SKILL.md). - Boundary markers: The instructions do not define specific delimiters (like XML tags or triple backticks) for wrapping user input during artifact generation.
- Capability inventory: The skill generates text-based artifacts including CSV, SQL DDL, and JSON Schema (referenced in
references/build-artifacts.md). It does not have access to shell execution or network tools (tools: []). - Sanitization: The skill includes logic to reject ambiguous user replies (e.g., 'yes', 'maybe') and enforces the use of 'Unknown' or 'Untested' for missing data to prevent fabrication. Given the primary purpose is data processing and no dangerous capabilities are present, the risk is negligible.
Audit Metadata