har-derived-api-client

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs extracting and re-sending captured cookies/Authorization headers from HARs (and har_to_client prints those replay hints), which requires the LLM/agent to handle and potentially output secret token values verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime ingests outsider-authored free text indirectly via the HAR it captures from a CDP-connected browser’s network traffic (e.g., response bodies and request payload text in har_capture_cdp.py / har_capture.py), then processes that text in har_to_client.py.