recoup-internal-consulting-invoice-generator
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it is designed to ingest and process external, potentially untrusted data from project files.
- Ingestion points: The agent is instructed to read Statement of Work (SOW) documents, payment structure definitions at
library/pricing/payment-structures.md, and workspace-specific configurations likeDESIGN.mdandAGENTS.mdinSKILL.md. - Boundary markers: The instructions lack explicit delimiters or safety headers to differentiate between system instructions and data content from the ingested files.
- Capability inventory: The skill possesses the capability to read from the local workspace and write generated financial logs to the
business/finance/directory as described inSKILL.md. - Sanitization: There is no evidence of validation, escaping, or sanitization protocols for the content extracted from the Statement of Work or other referenced documents.
Audit Metadata