gsd-workspace

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts arbitrary user input through the $ARGUMENTS variable and propagates it to internal workflow files.\n- Ingestion points: $ARGUMENTS parsed in SKILL.md.\n- Boundary markers: Absent; user arguments are not enclosed in delimiters or marked as untrusted before being passed to sub-workflows.\n- Capability inventory: The skill has access to Bash, Read, and Write tools (SKILL.md).\n- Sanitization: No sanitization or validation logic for arguments is present in the skill's processing block.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 12:43 AM
Security Audit — agent-trust-hub — gsd-workspace