casualize-names

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it interpolates untrusted data from Google Sheets into prompts for the Anthropic API.
  • Ingestion points: Data is ingested from the 'first_name', 'company_name', and 'city' columns of a user-provided spreadsheet in scripts/casualize_batch.py and the individual field scripts.
  • Boundary markers: While data is structured as JSON or numbered lists, the prompt lacks explicit delimiters or system instructions to ignore embedded commands within the data cells.
  • Capability inventory: The skill possesses the capability to perform network operations to the Anthropic API and write operations back to the Google Sheet via gspread.
  • Sanitization: No input validation, escaping, or filtering is applied to the spreadsheet content before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 02:18 PM