paseo-help
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the paseo CLI to run diagnostic commands like 'paseo --version' and 'paseo daemon status --json' to collect system information for troubleshooting.
- [EXTERNAL_DOWNLOADS]: Fetches product documentation and a documentation index ('llms.txt') from the vendor's official domain ('paseo.sh').
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data, creating a potential surface for indirect prompt injection.
- Ingestion points: Processes user input via the $ARGUMENTS variable and external documentation content fetched from 'https://paseo.sh/' (SKILL.md).
- Boundary markers: The skill uses textual labels (e.g., 'User's question:') to separate input but lacks formal structured delimiters or explicit instructions to ignore embedded commands.
- Capability inventory: The agent can execute shell commands (paseo CLI) and read local configuration and log files (SKILL.md).
- Sanitization: While no automated filtering is mentioned, the instructions explicitly command the agent to redact credentials, tokens, and passwords from log files before providing them to the user.
Audit Metadata