vuln-tracker

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill largely uses the secure gh CLI, but it includes an explicit curl fallback that sends a GitHub PAT in an Authorization header on the command line (Authorization: Bearer $GH_GLOBAL / {GH_GLOBAL}), which is an insecure pattern that could require embedding/verbatim use of secret values in generated commands.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s required runtime workflow “vuln-tracker” reads outsider-authored GitHub PR content by using gh pr view (Arm A step A3) and also ingests PR comments/review text and advisory data via gh api (Arm A/B), before notifying.