unseen-servant
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data from various sources to perform maintenance and triage, which presents a surface for indirect instructions to influence agent behavior.
- Ingestion points: The skill is explicitly intended to process email inboxes, git branches, and system logs as described in SKILL.md.
- Boundary markers: The procedure mandates the definition of 'no-touch boundaries' and requires an 'explicit confirmation' gate before taking actions that touch real system boundaries.
- Capability inventory: The skill utilizes the standard toolset for file modification, git branch management, and system automation.
- Sanitization: There are no specific instructions for escaping or sanitizing content from the external sources before processing.
Audit Metadata