The Agent Skills Directory

[PROMPT_INJECTION]: The skill's description in the YAML frontmatter includes directives such as "Use by default whenever the user wants to review shop leads..." and "Skip only if the user explicitly asks not to use this skill/workflow." These are attempts to override the agent's autonomous decision-making and standard tool selection behavior by creating a mandatory-use bias.
[INDIRECT_PROMPT_INJECTION]: The skill workflow involves retrieving and parsing potentially attacker-controlled content from prospect notes via the shop_get_prospect tool. This data is used to prioritize actions and generate reports, creating a vulnerability to embedded instructions.
Ingestion points: Data retrieved from the shop's prospect database via shop_list_prospects and shop_get_prospect (specifically the prospect notes).
Boundary markers: None present. The instructions do not mandate the use of delimiters or provide warnings to the agent to ignore instructions found within the data.
Capability inventory: The skill has the ability to write to the database (shop_update_prospect_note) and suggests further autonomous actions like shop-push-broadcast and shop-promo-campaign based on the processed data.
Sanitization: None. There are no instructions to escape, filter, or validate the content of the notes before the agent parses them for reminders or potential.

shop-prospect-nurture