debugging-master

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • Prompt Injection (SAFE): The skill instructions focus on defining a debugging persona and do not contain attempts to override agent behavior, extract system prompts, or bypass safety filters.
  • Data Exposure & Exfiltration (SAFE): No hardcoded credentials, sensitive file paths (like ~/.ssh or .env), or unauthorized network operations are present. Examples in patterns.md use local git commands and logging for educational purposes.
  • Unverifiable Dependencies & Remote Code Execution (SAFE): The skill does not reference external packages or scripts. There are no remote code execution patterns like piped shell commands (curl|bash).
  • Indirect Prompt Injection (SAFE): While the skill's purpose is to analyze user-provided code, it lacks the tools or capabilities (such as file-write or network-send) that would be required for an indirect prompt injection to be exploited. It behaves purely as a knowledge base and linter rule set.
  • Obfuscation (SAFE): All content is provided in clear text markdown and standard regex patterns with no encoding or hidden characters.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:12 PM