cicpa-company-query

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill instructs the agent to capture full cookie values (from Playwright or user paste) and embed that exact JSON into shell commands like python3 $SCRIPT --save-cookies '<...>', which requires the LLM to handle and output secret cookie values verbatim.

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

• Third-party content exposure detected (low risk: 0.10). SKILL.md/runtime path: the workflow uses user-supplied free text only as “company name / org_id” inputs to functions like search_company() / get_company_detail() (e.g., args.detail, args.names, args.file) which are then sent to first-party APIs; there is no workflow that ingests outsider-authored free text (e.g., email bodies, social feeds, GitHub/Jira issue text) without first selecting specific items.