tool-mastery
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill documents the use of remote scripts that are piped directly to the shell for execution.
- Evidence:
curl -L https://github.com/carlospolop/PEASS-ng/releases/latest/download/linpeas.sh | shinpayloads.md. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted target data and interpolates it into shell commands without sanitization.
- Ingestion points: Target variables like
TARGETanddomainare used throughoutpayloads.md,test-cases.md, and the guide files. - Boundary markers: Absent. The skill does not use delimiters to isolate untrusted input within command strings.
- Capability inventory: The skill utilizes the
Bashtool to execute a wide variety of powerful security and networking tools. - Sanitization: Absent. The skill does not perform any validation, escaping, or filtering of target variables before they are passed to the shell.
- [COMMAND_EXECUTION]: The skill is designed to enable the agent to run a large arsenal of security tools through the
Bashtool, representing a high-privilege operational capability. - [EXTERNAL_DOWNLOADS]: The skill documentation includes references for fetching security tools and configuration files from external GitHub repositories.
Audit Metadata