The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration specifies the installation of the wacli binary from external sources, specifically steipete/tap/wacli via Homebrew and github.com/steipete/wacli/cmd/wacli@latest using the Go toolchain.
[COMMAND_EXECUTION]: The skill's primary functionality is implemented through the execution of shell commands using the wacli CLI, which manages authentication, message history, and data transmission.
[DATA_EXFILTRATION]: The skill has the capability to read sensitive WhatsApp chat history and send local system files (e.g., wacli send file --file /path/agenda.pdf). This intended behavior facilitates the movement of local data to an external service.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. (1) Ingestion points: Untrusted data is ingested into the agent context from WhatsApp chat logs retrieved using wacli messages search or wacli chats list in SKILL.md. (2) Boundary markers: There are no markers or instructions defined to prevent the agent from executing commands that might be embedded in the retrieved WhatsApp messages. (3) Capability inventory: The skill enables the agent to send text and files to any WhatsApp recipient. (4) Sanitization: No sanitization or validation logic is present to filter incoming message content before it is processed.

wacli