supabase-dev-workflow

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (HIGH): The skill utilizes high-privilege tools including execute_sql and Bash(supabase:*) to manage database schemas and local development environments. These tools allow the agent to run arbitrary commands on the host system and database instance.
  • [PROMPT_INJECTION] (HIGH): Indirect Prompt Injection Risk. The agent reads and executes content from SQL files in the supabase/schemas/ directory. An attacker could inject malicious instructions into these files to achieve arbitrary code execution or data exfiltration when the agent processes them. Ingestion points: SQL files in supabase/schemas/, ENTITIES.md. Boundary markers: None. Capability inventory: execute_sql, Bash(supabase:*). Sanitization: None.
  • [SAFE] (INFO): The provided SQL patterns correctly use SECURITY INVOKER by default and SET search_path = '' to mitigate common PostgreSQL security risks like search path injection.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 06:21 AM