supabase-audit-auth-config
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Supabase API endpoints (e.g., /auth/v1/signup, /auth/v1/settings) and uses the responses to generate audit reports and update context files.
- Ingestion points: Data enters the agent's context through shell command outputs (curl) from the Supabase Auth endpoints defined in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings for the agent when processing these API responses.
- Capability inventory: The skill has the capability to write to multiple local files (.sb-pentest-context.json, .sb-pentest-audit.log, and files within .sb-pentest-evidence/) and execute shell commands via curl.
- Sanitization: There is no mention of sanitization or validation of the remote API response before it is processed or written to disk.
- [COMMAND_EXECUTION]: The skill uses the
curlcommand to interact with remote Supabase Auth endpoints. While this is the intended purpose of the audit tool, the use of user-supplied variables ($SUPABASE_URL and $ANON_KEY) in shell commands is a standard pattern for such tasks.
Audit Metadata