spreadsheet-workbook-forensics
Pass
Audited by Gen Agent Trust Hub on Jun 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The instructions direct the agent to write and run Python scripts to process spreadsheet data, which is the primary intended function of the skill.
- [REMOTE_CODE_EXECUTION]: The skill relies on the dynamic generation and execution of Python code (utilizing well-known libraries like
openpyxlandpandas) to manage workbook forensics and verification. - [PROMPT_INJECTION]: This skill presents an indirect prompt injection surface because it processes external spreadsheet data while retaining the capability to execute code and write to the file system. Analysis of surface (Category 8): 1. Ingestion points: Data is loaded from external workbooks specified by
INPUT_PATH(SKILL.md). 2. Boundary markers: None; the agent is not instructed to treat cell content as untrusted data or ignore embedded instructions. 3. Capability inventory: The agent has local script execution and file system read/write permissions via Python (SKILL.md). 4. Sanitization: Content normalization is mentioned for data consistency (e.g., whitespace collapsing, casefolding), but no security-focused sanitization of external data is present.
Audit Metadata