The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill executes 'npx -y @claude-flow/cli@latest', which downloads and runs code from the NPM registry at runtime. This introduces a supply chain risk, especially as the tool is used to handle sensitive session credentials.
[REMOTE_CODE_EXECUTION]: The use of 'npx' to fetch and execute a remote CLI tool without version pinning or integrity checks constitutes remote code execution from an untrusted source.
[DATA_EXFILTRATION]: The skill extracts raw browser cookies using 'document.cookie' and passes them to an external CLI tool ('memory store'). While the skill mentions an 'AIDefence' sanitization step, session tokens are sensitive data that are being transmitted to and stored in an external system.
[DYNAMIC_EXECUTION]: The skill uses the 'browser_eval' tool to execute JavaScript code ('document.cookie') within the browser context to retrieve session data.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection as it interacts with external web content. 1. Ingestion points: Web page content loaded during the login flow in SKILL.md. 2. Boundary markers: None present. 3. Capability inventory: Shell execution (Bash), browser script execution (browser_eval), and file writing (Write). 4. Sanitization: Uses 'aidefence_scan' for secrets/PII but not for preventing prompt injection.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute system commands for storing credentials in a remote namespace.

browser-login