skills/jeremylongshore/claude-code-plugins-plus-skills/managing-autonomous-development/Gen Agent Trust Hub
managing-autonomous-development
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to interact with a CLI utility namedsugar. It specifies several commands including/sugar-status,/sugar-review,/sugar-task, and/sugar-run. These commands are standard for the described task management and autonomous development workflow. - [EXTERNAL_DOWNLOADS]: The skill references external resources for documentation and source code hosted on
github.com/roboticforce/sugarandroboticforce.io. These are used neutrally as reference materials for the user. - [PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection as it processes user-supplied strings in the
<description>field of the/sugar-taskcommand. - Ingestion points: User-provided task descriptions entered via the
/sugar-taskcommand inSKILL.md. - Boundary markers: Absent; descriptions are interpolated into command arguments without explicit delimiters.
- Capability inventory: The skill has broad capabilities including
Bash(cmd:*),Read,Write, andEdit(specified in theallowed-toolsfrontmatter). - Sanitization: No explicit sanitization or validation of the input description is described in the instructions.
Audit Metadata