prompt-engineering

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill provides educational content on identifying and preventing prompt injection. It includes a sanitize_user_input function and instructions on using delimiters to isolate user input from system instructions, which are security best practices.
  • [DATA_EXFILTRATION]: No patterns of unauthorized data access or transmission to external domains were identified. The code snippets demonstrate standard API interactions with the Anthropic client.
  • [REMOTE_CODE_EXECUTION]: The skill does not contain any remote script downloads, piped execution, or unsafe dynamic evaluation of untrusted strings.
  • [COMMAND_EXECUTION]: No shell command execution or unauthorized subprocess spawning patterns were found.
  • [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or sensitive credentials are present in the provided files.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 05:21 PM
Security Audit — agent-trust-hub — prompt-engineering