The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external websites (such as competitor ad libraries and landing pages) using the agent-browser tool, as well as local brand context files. This data is processed to generate marketing deliverables like campaign briefs and ad copy.
Ingestion points: External website content is fetched via agent-browser in references/competitive-research.md, and local files such as brand-context.md are read as described in SKILL.md.
Boundary markers: The instructions do not define explicit boundary markers or delimiters to separate ingested external data from the agent's core instructions.
Capability inventory: The skill has the ability to perform network operations and execute JavaScript on external sites via agent-browser, and it writes deliverables to the local .pawbytes/ directory.
Sanitization: There is no description of sanitizing, filtering, or escaping the external content before it is interpolated into the agent's context.
[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the agent-browser tool from a public registry or a trusted GitHub repository if it is not already present in the environment.
Evidence: The file references/shared-patterns.md suggests using npx skills add https://github.com/vercel-labs/agent-browser to install the required automation tool.
[COMMAND_EXECUTION]: The skill utilizes shell commands to perform browser automation, verify tool availability, and manage software packages.
Evidence: references/competitive-research.md and references/shared-patterns.md contain specific command lines for agent-browser, npx, npm, and local discovery scripts.

paw-mkt-paid-ads