The Agent Skills Directory

[PROMPT_INJECTION]: The skill retrieves untrusted content from external news providers and TradingView's HTML subpages which is then incorporated into the agent's context, creating a surface for indirect prompt injection attacks where malicious instructions could be embedded in the market news or site content.
Ingestion points: News headlines and bodies fetched in scripts/fetch_tradingview.py via the news_by_symbol and news_story functions, as well as data extracted from HTML scraping in fetch_subpage.
Boundary markers: Data is ingested into the agent context without explicit delimiters or instructions to the model to ignore potentially embedded commands.
Capability inventory: The skill is restricted to network operations using the requests library and local file system reads for its own assets. It does not exhibit arbitrary command execution or dynamic code evaluation capabilities.
Sanitization: While the script uses regular expressions to strip HTML tags from news content, it does not implement specific sanitization or filtering to prevent prompt injection.
[EXTERNAL_DOWNLOADS]: The script fetches data from several subdomains of tradingview.com (a well-known financial service) to provide market quotes, symbol search results, and news. These network operations are consistent with the skill's documented purpose of providing market analysis tools and target well-known infrastructure.

tradingview