prophet-adversarial-auditor

Pass

Audited by Gen Agent Trust Hub on Mar 21, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection surface. * Ingestion points: Historical market data and submission events enter the context via load_run_history as defined in the skill workflow. * Boundary markers: The skill does not explicitly define boundary markers or provide instructions to ignore embedded commands when processing ingested market data. * Capability inventory: The agent has network access to app.prophetmarket.ai and api.serendb.com (vendor resources) via urllib.request, and database write access via psycopg. * Sanitization: No explicit sanitization or filtering of the ingested external data is performed in the main runtime script before processing.
  • [DATA_EXFILTRATION]: The skill performs network requests to app.prophetmarket.ai (the target auditing service) and api.serendb.com (the storage infrastructure associated with the vendor). These operations are required for the skill's primary function and do not target unauthorized or suspicious domains.
  • [COMMAND_EXECUTION]: The script executes SQL DDL statements from a local file (serendb_schema.sql) to initialize the storage schema. This is standard setup behavior for establishing the necessary database tables in the user's provisioned storage instance.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 21, 2026, 02:44 AM