The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx -y @claude-flow/cli@latest to download and execute a command-line tool for template storage and retrieval. This package is part of the Claude Flow ecosystem used by the skill.
[COMMAND_EXECUTION]: Utilizes Bash to manage session logic, execute template retrieval, perform string manipulation for PII redaction, and store extracted data.
[PROMPT_INJECTION]: The skill ingests untrusted data from external web pages, creating an indirect prompt injection surface.
Ingestion points: Untrusted data enters the context via mcp__claude-flow__browser_snapshot and mcp__claude-flow__browser_eval (SKILL.md).
Boundary markers: The skill does not define specific delimiters for the extracted content, but it mandates the use of a safety scanner.
Capability inventory: The skill has access to Bash for command execution and Write for file system operations.
Sanitization: Implements mcp__claude-flow__aidefence_is_safe and mcp__claude-flow__aidefence_has_pii to scan and redact malicious or sensitive content before it is returned to the model.

browser-extract