The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources such as Google search results, Reddit threads, and Quora questions to perform market research and content ideation. This creates an attack surface where malicious instructions embedded in web content could influence the agent's behavior. Ingestion points: The references/content-research.md file utilizes the agent-browser tool to extract text from external URLs. Boundary markers: No specific delimiters are used to isolate untrusted web text within the prompt instructions. Capability inventory: The skill possesses file-write capabilities in SKILL.md and network access in references/content-research.md. Sanitization: The references/frameworks/ai-content-workflow.md prescribes a mandatory human review phase to check for hallucinations and maintain quality control.
[REMOTE_CODE_EXECUTION]: The skill facilitates the installation of the agent-browser tool from the official Vercel Labs GitHub repository. While this involves executing external code, the source is a verified and trusted organization, making this operation a safe dependency for the skill's functionality.

paw-mkt-content