sqlite-explorer

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill allows the execution of arbitrary SQL queries, including data modification and schema changes (DDL), on any file accessible to the agent. While intended for database exploration, this provides a powerful mechanism for file and data manipulation.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local database files without boundary markers or sanitization, creating an attack surface for indirect prompt injection.
  • Ingestion points: The runQuery, showSchema, and listTables functions in scripts/sqlite-explorer.ts read content from external SQLite files and display them in the agent's context.
  • Boundary markers: Absent. There are no instructions or technical controls to prevent the agent from interpreting data within the database as new instructions.
  • Capability inventory: The agent has the ability to execute destructive SQL queries (DELETE, DROP, etc.) and read/write to the file system.
  • Sanitization: Absent. The script prints the database content directly to the console.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 10:30 AM
Security Audit — agent-trust-hub — sqlite-explorer