The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of sub-process spawning to run local maintenance and SEO scripts such as backlink/scripts/semrush-keyword.mjs, rankup/scripts/seo-webcafe.mjs, and rankup/scripts/gt.py. It also invokes external CLI tools like yt-dlp and opencli to gather data from YouTube, Reddit, and X (Twitter).
[EXTERNAL_DOWNLOADS]: The automation logic involves fetching and monitoring content from numerous external domains, including gaming platforms (itch.io, Poki, CrazyGames, Steam), social media sites, and professional SEO tools (Semrush, Similarweb). The checkUrl function in scripts/game-opportunity.mjs performs direct HTTP fetches on discovered game URLs.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from public forums and social media to discover new keywords and game titles.
Ingestion points: The socialRadar function in scripts/game-opportunity.mjs ingests text and metadata from Reddit search results, YouTube video info, and X posts.
Boundary markers: While the script parses these as structured JSON objects, it does not apply specific delimiters or instructions to prevent the agent from being influenced by malicious instructions embedded in the titles or descriptions of the scraped content.
Capability inventory: The agent has the capability to write research results to the .rankup/ directory and execute further command-line tools based on the extracted keywords.
Sanitization: The script performs basic HTML entity decoding and whitespace normalization but lacks semantic filtering for potentially malicious prompt content.

game-opportunity