agent-devsecops-engineer
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses an unconstrained input placeholder (
$ARGUMENTS) at the end of the instructions. This creates a surface where external data passed to the agent could contain instructions that influence the agent's behavior. However, the skill does not currently possess dangerous capabilities like direct file writing or network exfiltration, limiting the risk. - Ingestion points:
SKILL.md(via$ARGUMENTSplaceholder). - Boundary markers: None present around the input placeholder.
- Capability inventory: None (the skill provides instructions and examples but no executable scripts).
- Sanitization: None observed for the input placeholder.
Audit Metadata