harden
Pass
Audited by Gen Agent Trust Hub on Aug 16, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill's primary function is to guide an AI agent through the process of establishing a robust development feedback loop. No malicious activities, data exfiltration, or obfuscation were found.
- [COMMAND_EXECUTION]: As part of its intended dev-loop setup, the skill instructs the agent to identify, define, and run verification commands (e.g.,
bun run verify). These actions are transparent and align with the skill's stated purpose of project hardening. - [INDIRECT_PROMPT_INJECTION]: The skill includes a surface for indirect instructions through user-supplied arguments used to specify which tools to configure. \n
- Ingestion points: The
$ARGUMENTSvariable inSKILL.mdcaptures user input. \n - Boundary markers: None are explicitly defined to isolate the user-provided arguments within the prompt. \n
- Capability inventory: The agent can modify project configuration files and execute local validation scripts as defined by the user and the project's environment. \n
- Sanitization: No input sanitization or validation logic is present in the skill instructions.
Audit Metadata