ask-agy
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to invoke the
agyCLI tool via Bash commands to perform coding tasks. - [PROMPT_INJECTION]: The skill promotes the use of the
--dangerously-skip-permissionsflag in its examples. This flag explicitly disables the confirmation prompts that theagytool would otherwise present before performing potentially sensitive or destructive actions (such as overwriting files or executing code), significantly reducing the security controls of the underlying system. - [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it passes user-controlled content directly to the CLI's prompt parameter.
- Ingestion points: Content passed to the
-por--promptflag inSKILL.mdis processed by the external CLI tool. - Boundary markers: There are no delimiters or instructions to ignore embedded commands within the prompt input, allowing untrusted data to potentially control the CLI's behavior.
- Capability inventory: The
agytool has the capability to modify local files and generate code based on its input instructions. - Sanitization: The skill does not implement sanitization or validation of the text before it is passed to the command line.
Audit Metadata