The Agent Skills Directory

[DATA_EXPOSURE]: The skill reads local project context files like .agents/product-marketing.md to refine its research. This is a standard pattern for providing domain context to an agent and does not target sensitive system or credential files.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content scraped from external competitor websites via the firecrawl_scrape tool. This represents an indirect prompt injection surface where an attacker-controlled website could include malicious instructions meant to subvert the agent's synthesis logic.
Ingestion points: Raw markdown and JSON data retrieved from external URLs using Firecrawl and DataForSEO.
Boundary markers: The skill does not explicitly instruct the agent to use delimiters or ignore embedded instructions when reading back saved raw data during the synthesis phase.
Capability inventory: The skill has the capability to write research results to the local competitor-profiles/ directory and perform network requests via defined MCP tools.
Sanitization: There is no mention of sanitizing or filtering the scraped content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill utilizes Firecrawl and DataForSEO, which are well-known technology services for web scraping and search engine data. These tools are used for their intended purpose of gathering competitive intelligence.

competitor-profiling