skills/nanocoai/nanoclaw/add-whatsapp/Gen Agent Trust Hub

add-whatsapp

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands via nc:run to manage configuration files (sed, grep), install dependencies (pnpm), and handle service lifecycles (systemctl, launchctl, bash).
  • [EXTERNAL_DOWNLOADS]: Fetches external dependencies from the NPM registry and copies files from a remote git branch (channels) to integrate the WhatsApp adapter.
  • [INDIRECT_PROMPT_INJECTION]: As the skill enables the agent to read and respond to WhatsApp messages, it introduces a surface for indirect prompt injection from external users.
  • Ingestion points: Reads WhatsApp messages (DMs and groups) via the Baileys adapter.
  • Boundary markers: The instructions distinguish between shared and dedicated modes, advising name-based engagement to prevent accidental triggers.
  • Capability inventory: Includes shell command execution, environment variable modification, and network operations.
  • Sanitization: Employs regex-escaped agent names and word-boundary guards for engagement patterns.
  • [DYNAMIC_EXECUTION]: Utilizes tsx and node -e to execute logic at runtime, including a local HTTP server in scripts/wa-qr-browser.ts to display authentication QR codes in a web browser.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 06:59 PM