The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes content from external, untrusted sources (competitor websites).
Ingestion points: Data enters the agent's context through firecrawl_map, firecrawl_scrape, and firecrawl_search operations defined in SKILL.md.
Boundary markers: The instructions lack explicit boundary markers (e.g., XML tags) or specific directives to the agent to ignore or neutralize instructions embedded within the scraped website content.
Capability inventory: The skill has the capability to write research profiles and raw data to the local filesystem (competitor-profiles/ directory) and read project context files (.agents/product-marketing.md).
Sanitization: There is no mention of sanitizing, escaping, or filtering the scraped content before it is synthesized into final reports.
[COMMAND_EXECUTION]: The skill uses various MCP tools for its operations. While it performs network requests to fetch data, these are directed at well-known services (Firecrawl, DataForSEO) and target domains specified by the user for research purposes, which is consistent with the skill's primary function.

competitor-profiling