skills/vllnt/skills/manage-issues/Gen Agent Trust Hub

manage-issues

Pass

Audited by Gen Agent Trust Hub on Nov 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues or malicious patterns detected. The skill focuses on repository maintenance and issue triage using standard GitHub interactions. It contains explicit safety guidelines, such as read-only defaults for analysis and non-destructive issue closure.
  • [INDIRECT_PROMPT_INJECTION]: The skill inherently processes untrusted external data from GitHub issue backlogs and repository documentation. This represents a potential indirect prompt injection surface. However, the skill implements several mitigating controls: it explicitly limits the agent to triage rather than implementation ('without implementing its issues'), mandates a quality gate to verify intent ('understand its outcome without inventing intent'), and requires a formal validation protocol (quality-validation.md) for any authorized writes. The workflow emphasizes evidence-based decisions and human-in-the-loop authorization, reducing the risk of autonomous exploitation by malicious issue content.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 6, 2026, 04:54 PM