openapi-analyzer

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

• Third-party content exposure detected (high risk: 0.75). The skill fetches an OpenAPI JSON spec from an external/internal MikoPBX URL at runtime, and the analyzer loads it into memory and parses it as JSON (readable text) via curl ... > /tmp/mikopbx_openapi.json then json.load(f) in MikoPBXOpenAPIAnalyzer.__init__, so any free-text fields inside that spec (e.g., summary/description) are outsider-authored content.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)