The Agent Skills Directory

Indirect Prompt Injection Surface: The skill is designed to fetch and process content from external, user-provided URLs and web search results to research business models. This ingestion of untrusted data presents a potential surface where malicious instructions embedded in external websites could attempt to influence the agent's behavior.
Ingestion points: Data enters the agent's context through WebFetch operations on arbitrary URLs provided by the user and through WebSearch results, as detailed in references/company-researcher.md.
Boundary markers: The instructions do not currently define explicit delimiters or 'ignore' instructions for the content retrieved from external sources before it is analyzed by the model.
Capability inventory: The skill utilizes WebFetch, WebSearch, Grep, and Glob to retrieve data from both the internet and the local file system.
Sanitization: There are no explicit sanitization or filtering steps mentioned for the fetched content before it is used for business model classification.
Codebase Information Gathering: The skill automatically performs scans of the local project using tools like Grep and Glob to identify existing Stripe integration patterns.
Pattern: It searches for sensitive keywords such as connected_account, application_fee_amount, and stripe_account to understand the current technical context.
Context: While this is a functional requirement for providing accurate integration advice, users should be aware that the agent will autonomously read parts of the codebase to extract this logic.

connect-recommend