record
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the creation and execution of shell commands for terminal recording using the
asciinemaCLI utility. - [DATA_EXFILTRATION]: The skill description includes a
--backupflag for streaming session recordings to GitHub. This involves the transmission of terminal session data, which could potentially include sensitive information, to a well-known external service. - [EXTERNAL_DOWNLOADS]: The documentation references the installation of the
asciinemadependency through the Homebrew package manager (brew install asciinema). - [PROMPT_INJECTION]: The skill processes user-supplied arguments to build command lines, which presents a surface for indirect prompt injection.
- Ingestion points: Arguments such as
file,title, andidle-time-limitare sourced directly from user input (file:SKILL.md). - Boundary markers: No markers or instructions are provided to the agent to delimit or ignore instructions within these user-supplied fields.
- Capability inventory: The skill utilizes the
Bashtool for command execution (file:SKILL.md). - Sanitization: There is no evidence of input validation, escaping, or sanitization logic to prevent command injection or manipulation via the provided arguments.
Audit Metadata