The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation provides instructions to install and run the 'octagon-mcp' tool using the Node package runner (npx). This tool is provided by the skill author (OctagonAI) and is part of the required infrastructure for the skill to function.
[EXTERNAL_DOWNLOADS]: Installation steps for prerequisites include a command to download and execute the Homebrew installation script from its official GitHub repository ('raw.githubusercontent.com/Homebrew/install/HEAD/install.sh'). This targets a well-known and trusted service.
[COMMAND_EXECUTION]: The setup guide includes commands for users to configure their environment, such as setting the 'OCTAGON_API_KEY' and launching the MCP server. These are standard configuration procedures for the platform.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external financial data (analyst ratings and institutional sentiment) which could theoretically contain malicious instructions if the upstream data source were compromised.
Ingestion points: Analyst ratings, institution names, and historical rating data retrieved from the Octagon API via the 'octagon-agent' tool.
Boundary markers: None explicitly defined in the instructions to the agent.
Capability inventory: The skill uses specific MCP tools ('octagon-agent', 'octagon-scraper-agent') for market intelligence gathering.
Sanitization: The skill assumes the integrity of the financial data provided by the Octagon platform.

stock-grades