The Agent Skills Directory

[SAFE]: The skill implements a structured multi-agent workflow for marketing strategy. It adheres to standard practices for agent coordination and uses platform tools for their intended purposes without malicious intent.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from marketing research files. However, the impact is limited to the generation of marketing strategy text rather than the execution of system commands or data exfiltration. \n
Ingestion points: .agents/mkt/icp-research.md, .agents/mkt/content-research.md, .agents/product-context.md, and .agents/solution-design.md.\n
Boundary markers: Absent; the orchestrator does not specify delimiters for external content.\n
Capability inventory: Access to Read, Grep, Glob, Bash, WebSearch, and WebFetch tools.\n
Sanitization: No explicit sanitization or validation of the ingested markdown content is present.

imc-plan