token-launch

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • SAFE (SAFE): No security issues were detected. The skill consists of instructional metadata and a series of regex-based patterns used for identifying common vulnerabilities in smart contracts (e.g., reentrancy, hardcoded decimals, unsafe transfers).
  • Indirect Prompt Injection (SAFE): While the skill is designed to ingest and analyze user-provided code (untrusted data), it lacks high-risk capabilities such as arbitrary command execution, network access, or file-writing. The instructions explicitly direct the agent to prioritize the provided reference files over user instructions when conflicts arise, which serves as a protective grounding mechanism.
  • Metadata Analysis (SAFE): Metadata fields are accurate and descriptive of the skill's purpose. No deceptive content or injection markers found in the YAML frontmatter.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:25 PM