The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the agent-browser tool, which it fetches from a trusted GitHub repository. The installation process uses npx skills add to integrate the capability.
[COMMAND_EXECUTION]: The skill instructs the agent to run browser automation commands using the agent-browser CLI. These commands are dynamically constructed using variables such as {competitor-domain} to perform market research.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection attacks because it processes data from external sources. 1. Ingestion points: The skill fetches and reads the body text of external websites (competitor pages, review sites like G2/Capterra, and LinkedIn) through agent-browser. 2. Boundary markers: There are no explicit instructions to the agent to use delimiters or to disregard instructions embedded in the external content. 3. Capability inventory: The agent has capabilities for browser automation, capturing screenshots, and writing files to the local project structure. 4. Sanitization: No sanitization or validation logic is applied to the data retrieved from external sources before it is used to generate sales collateral.

marketing-sales