The Agent Skills Directory

[NO_CODE]: This skill consists of markdown logic and instructions for an AI agent and does not include any scripts, binaries, or configuration files.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process external, potentially untrusted data for transmission to mobile channels.
Ingestion points: External data sources including revenue reports, invoices, and server error logs as referenced in SKILL.md enter the agent context.
Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions embedded within the ingested data.
Capability inventory: The skill describes the use of tools like whatsapp_send_message and channel_search (referenced in SKILL.md) to interact with external messaging platforms.
Sanitization: There is no mention of sanitizing, escaping, or validating the content of the ingested reports or logs before they are processed by the agent.

business-messaging-automation