e2e-agent-browser

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill prompt embeds a plaintext password ("password123") and other credential-like values directly in example commands/scripts and even includes an eval that reads a stored token, which would force an LLM to output secrets verbatim if it reproduces those commands or results.

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

• Third-party content exposure detected (medium risk: 0.30). templates/runner.js and templates/test-suite.sh run-time ingest free text from the loaded web pages by calling agent-browser snapshot/title/text/wait, which can include attacker-authored content from the target site.