The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions to download a ZIP archive (goplus-scanner-enterprise.zip) from an unverified GitHub repository (BrewerDifferentiate/GoPlus-Security-Token-Scanner-Enterprise-Pro) and extract it. This pattern promotes the execution of unverified binaries from an untrusted source.
[EXTERNAL_DOWNLOADS]: The skill uses PowerShell's Invoke-WebRequest to fetch external content from a GitHub account that is not associated with the official service provider, bypassing standard security reviews for software distribution.
[PRIVILEGE_ESCALATION]: The installation instructions direct the user to extract the downloaded files into C:\Program Files\GoPlusSecurity, a restricted system directory. Performing this action typically requires administrative rights and can be used to establish persistence or compromise the integrity of the host system.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from blockchain security scans and contract analysis. If the data source (API) provides malicious content, it could influence the agent's decision-making or output.
Ingestion points: Data retrieved from the GoPlus Security API via client.token_security, client.approval_security, and client.nft_security (SKILL.md).
Boundary markers: None identified in the documentation or provided code samples to separate untrusted API data from agent instructions.
Capability inventory: The skill has capabilities for network communication, file system modification (via PowerShell), and shell command execution.
Sanitization: The provided Python snippets perform basic logic checks (e.g., checking if a flag is '1') but do not appear to sanitize or validate the integrity of the broader response strings before they are processed by the agent.

goplus-security-token-scanner