research-opportunity-investigator

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains no executable code or scripts. It primarily consists of instructional markdown and templates designed to guide an agent through a manual research process.
  • [PROMPT_INJECTION]: The skill uses directive language to enforce strict procedural constraints (e.g., "NEVER Proceed Without Explicit User Confirmation"). These are intended to improve accuracy and user oversight rather than bypass safety guardrails.
  • [DATA_EXFILTRATION]: While the skill involves searching the web for documentation and repositories, it does not access sensitive local files or hardcoded credentials. No unauthorized network transmission patterns were found.
  • [REMOTE_CODE_EXECUTION]: The skill does not include any commands for downloading external software or executing remote scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 02:52 AM
Security Audit — agent-trust-hub — research-opportunity-investigator