terminal-ops
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to guide an agent through executing terminal commands, inspecting git states, and performing repository modifications (fixes, commits, and pushes).
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it instructs the agent to ingest and process data from potentially untrusted external sources (repository files, CI failure logs, and git outputs) while maintaining active execution capabilities.
- Ingestion points: Repository files, git state, CI logs, and terminal command outputs.
- Boundary markers: The instructions do not specify any delimiters or warnings for the agent to ignore embedded instructions within the data being read.
- Capability inventory: The agent is authorized to run arbitrary commands, inspect file systems, and perform git operations (commit/push) based on the workflow.
- Sanitization: There are no explicit instructions for sanitizing or escaping the data ingested from the repository or environment before processing.
Audit Metadata