The Agent Skills Directory

[DATA_EXFILTRATION]: The script scripts/extract_ranked_keywords.py searches for and reads environment configuration files (.env, .env.local, .dataforseo.env) from the current directory, parent directories up to the git root, and the user's home directory (~). While intended for credential management, searching the home directory for generic environment files is a sensitive file access pattern.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to https://api.dataforseo.com to fetch ranking data. This is a well-known service and the behavior is central to the skill's purpose.
[INDIRECT_PROMPT_INJECTION]: The skill ingests keyword data and intent descriptions from an external API response which are then interpolated into a Markdown table, creating a potential surface for indirect prompt injection.
Ingestion points: target_url (CLI argument) and the JSON response from the DataForSEO API processed in scripts/extract_ranked_keywords.py.
Boundary markers: The to_markdown function wraps the keyword data in a Markdown table structure.
Capability inventory: The script performs network operations via urllib.request and reads system environment files.
Sanitization: Performs URL normalization and escapes pipe characters for Markdown, but lacks specific content validation to prevent instruction injection from the retrieved keyword data.

url-ranked-keywords-extractor