agent-whatsapp
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the
agent-messengerpackage from the NPM registry to provide theagent-whatsappCLI functionality. This package is the core tool used by the skill for its primary functions. - [DATA_EXPOSURE]: WhatsApp authentication state and session keys are stored locally in
~/.config/agent-messenger/. While the skill sets restrictive filesystem permissions (0600), the documentation notes that these keys are stored in plaintext and grant full access to the linked WhatsApp account. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external WhatsApp chat data, creating an attack surface for indirect prompt injection.
- Ingestion points: Untrusted message content is retrieved via the
agent-whatsapp message listcommand. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between message content and system instructions.
- Capability inventory: The agent has access to a restricted shell environment (
Bash(agent-whatsapp:*)) and file system access via standardRead/Writetools. - Sanitization: No sanitization or filtering logic is mentioned for processing the body of incoming messages.
Audit Metadata