The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to crawl external websites, which introduces an indirect prompt injection surface if downstream tools or prompts process the retrieved content without sanitization.
Ingestion Points: Content is ingested from external web URLs via the firecrawl crawl command and stored locally in .firecrawl/crawl.json.
Boundary Markers: Absent; there are no instructions or configurations defining boundaries for untrusted content.
Capability Inventory: The skill utilizes the Bash tool to run firecrawl and npx firecrawl-cli commands.
Sanitization: Absent; no parsing or validation mechanisms are defined within the skill to filter out malicious instructions from the crawled data.

firecrawl-crawl