extracting-filemaker-business-logic
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill references a specific local absolute file path in its examples:
/Users/anthonybyrnes/PycharmProjects/Python419/AugustServer_ddr/AugustServer.html. This exposes internal directory structures and the username of the author's development environment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and parse large (5-20MB) external HTML files (FileMaker DDR reports) which could be controlled by an attacker to include hidden instructions.
- Ingestion points: Large HTML files located at
{ProjectName}_ddr/{DatabaseName}.html. - Boundary markers: Absent; there are no instructions to use delimiters or 'ignore' commands when processing the external HTML content.
- Capability inventory: The skill encourages the agent to parse files using
lxmland generate/execute database commands usingcursor.executeor Python transformation logic based on the findings. - Sanitization: Absent; the skill does not recommend any sanitization or validation of the calculation formulas or script logic extracted from the external HTML reports before they are processed by the agent.
Audit Metadata