stitch-skills
Warn
Audited by Socket on Apr 22, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: The core UI-design and screen-to-code capabilities are coherent with the stated purpose, and the Stitch-related data flow appears proportionate. The main risk is transitive trust: this skill instructs the agent to install additional skills via npx, including from an unrelated third-party GitHub repo, which expands the execution surface beyond the publisher and raises supply-chain risk. No strong evidence of credential theft or clearly malicious exfiltration is present, but the install footprint is broader than necessary.
Confidence: 83%Severity: 62%
Audit Metadata