reusable-commands
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the creation of configuration files that utilize shell command execution (using OpenCode's
!commandsyntax) and file system access. This functionality is intended for legitimate developer automation, such as running test suites or checking git logs, but provides a mechanism for local command execution. - [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface by relying on user-provided descriptions and instructions to generate persistent executable command templates. An attacker could potentially use this to trick the agent into creating malicious commands on the user's filesystem.
- Ingestion points: User input for 'Command Name', 'Description', and 'Prompt Body' (SKILL.md).
- Boundary markers: Absent; the skill does not implement delimiters or safety warnings to wrap generated prompt content.
- Capability inventory: File system write operations to create commands in
.opencode/commands/and.github/prompts/(SKILL.md). - Sanitization: Absent; the agent is not instructed to validate or sanitize the shell commands or file paths provided in the user's prompt body.
Audit Metadata