fetch
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses the
$ARGUMENTSvariable to dynamically construct file paths (e.g.,skills/designs/$ARGUMENTS.mdandskills/brands/$ARGUMENTS/SKILL.md) and then displays the content of these files to the agent. This creates a vulnerability surface where a user could potentially attempt path traversal or where a design file could contain malicious instructions that influence the agent's next actions. - Ingestion points:
SKILL.md(logic to read and display files based on user input). - Boundary markers: None present; the skill does not use delimiters or instructions to ignore embedded commands within the fetched files.
- Capability inventory: File system read access (implied by the instruction to 'check' and 'display' files).
- Sanitization: No explicit sanitization or validation of the
$ARGUMENTSvariable is defined to prevent path traversal (e.g., using../to access files outside the intended directory).
Audit Metadata