accessibility-wcag

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The prompt contains an explicit "Companion check" that instructs the agent to run filesystem commands (ls ~/.claude/skills/ ...) and offer installations (npx skills add ...) which are side-effectful and unrelated to the skill's stated accessibility purpose, so it is an out-of-scope, potentially sensitive instruction.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs auditing arbitrary public sites (e.g., "Audit accessibility with axe-core and Lighthouse" with commands like npx lighthouse https://your-site.com and npx axe https://your-site.com), which entails fetching and analyzing untrusted third‑party web content that can influence the agent's findings and follow-up actions.