agent-registry
Fail
Audited by Snyk on Jul 9, 2026
Risk Level: CRITICAL
Full Analysis
CRITICAL E005: Suspicious download URL detected in skill instructions.
- Suspicious download URL detected (high risk: 1.00). These URLs include a raw GitHub-hosted install.sh that the docs instruct users to curl | bash (high-risk), and GitHub repo links to an individual/unknown account that host that install script — executing remote shell scripts from personal repos is a common malware delivery vector.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.95). The required runtime workflow for this skill (polling arena turns and generating replies) necessarily ingests outsider-authored free-form chat text from other participants/rooms via the arena API into the agent’s LLM context for “Parse → Think → respond”.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The skill contains an install one-liner that runs "curl -fsSL https://raw.githubusercontent.com/Solizardking/solana-clawd/newnew/cheshire-terminal/arena/install.sh | bash", which fetches remote shell code at runtime and pipes it to bash (remote code execution during setup).
MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).
- Direct money access detected (high risk: 1.00). The skill explicitly provides on-chain crypto payment capabilities: it requires a SOLANA_PRIVATE_KEY, shows how to construct/sign/send Solana transactions programmatically (createTransferInstruction, sendTransaction) and via CLI (spl-token transfer), references the CLAWD SPL token mint and Helius RPC, and describes using txSignatures as proof-of-payment. These are specific crypto/Blockchain wallet and transaction instructions — i.e., direct financial execution.
Issues (4)
E005
CRITICALSuspicious download URL detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
W009
MEDIUMDirect money access capability detected (payment gateways, crypto, banking).
Audit Metadata