The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the agent-browser tool from Vercel Labs' GitHub repository. This is an official source from a trusted organization intended to enable automated web research and auditing capabilities.
[COMMAND_EXECUTION]: Includes setup commands for installing dependencies and preparing the browser environment using npm install and npx playwright install. These commands are part of the standard deployment process for the tools required by the skill.
[PROMPT_INJECTION]: The skill is configured to audit live website URLs and inspect local codebases to provide marketing strategy recommendations. This processing of untrusted external content represents a surface for indirect prompt injection.
Ingestion points: Analyzes user-supplied URLs and repository source files as part of the context routing and audit workflows described in references/context-router.md.
Boundary markers: Absent; there are no specific instructions or delimiters provided to the agent to treat the ingested external content as non-executable data or to ignore embedded instructions.
Capability inventory: The agent can read and write files within the project directory, manage brand-specific workspaces, and coordinate multiple specialist agents.
Sanitization: Absent; the skill does not define any validation or sanitization steps for the data retrieved from external URLs or files before it is processed.

bmad-agent-marketing-agency