The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by incorporating untrusted data from local configuration files into its execution flow.
Ingestion points: Information such as target_states, preferred_makes, and price ranges is read from the marketcheck-profile.md file.
Boundary markers: The skill does not define delimiters or provide instructions for the agent to ignore or isolate potentially malicious instructions that could be embedded within the project memory file.
Capability inventory: The skill utilizes Marketcheck API tools (mcp__marketcheck__search_active_cars, mcp__marketcheck__get_sold_summary) and has the capability to spawn sub-agents via the Agent tool, which could be exploited if malicious data is ingested.
Sanitization: Input validation and escaping mechanisms are absent; variables extracted from the profile are interpolated directly into prompts and tool parameters.

territory-dashboard