tdp-plugin-backup
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input (the plugin name) to construct filesystem paths for destructive operations (deletion) and recursive copying. Without explicit instructions for input sanitization, this creates a path traversal surface.
- Ingestion points: User request containing the plugin name (SKILL.md).
- Boundary markers: Absent; no instructions are provided to the agent to treat the plugin name as untrusted or to validate it against a specific character set.
- Capability inventory: The skill possesses the ability to delete contents of directories and perform recursive file copies (SKILL.md).
- Sanitization: Absent; there is no requirement to escape or validate the input. A malicious user could provide a plugin name like
../../.sshto redirect the backup operation to sensitive system directories.
Audit Metadata