query-budget-enforcer
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted SQL query data and possesses the capability to write reports to the local file system, creating an attack surface where malicious input could influence file content.
- Ingestion points:
SKILL.mdspecifies that the agent should review SQL queries (provided by a user or tool) to perform budget and performance assessments. - Boundary markers: The skill instructions do not utilize boundary markers or explicit prompts to ignore natural language instructions that might be embedded within the SQL code (e.g., inside comments).
- Capability inventory: The skill explicitly writes both Markdown and JSON reports to
docs/database-report/query-budget-report.mdanddocs/database-report/query-budget-output.jsonas described inSKILL.md. - Sanitization: No sanitization, validation, or escaping logic is defined to protect the output reports from malicious strings contained within the analyzed SQL queries.
Audit Metadata