skills/knocklabs/skills/knock-cli/Gen Agent Trust Hub

knock-cli

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the Knock CLI using official vendor packages, specifically @knocklabs/cli via npm and the knocklabs/tap/knock Homebrew tap. These are official distribution points for the vendor's tooling.
  • [INDIRECT_PROMPT_INJECTION]: The documentation describes using Liquid templating to interpolate dynamic data from trigger payloads and user properties into notification templates. This defines a standard attack surface for indirect prompt injection common to notification platforms.
  • Ingestion points: Trigger payload data (referenced via the data. namespace) and recipient properties (referenced via recipient.) used in workflows, templates, and partials.
  • Boundary markers: Standard Liquid template delimiters are used; no additional boundary markers or explicit safety instructions regarding untrusted content are documented.
  • Capability inventory: CLI tools include commands to fetch, modify, and push templates (knock pull, knock push) and trigger testing runs (knock workflow run).
  • Sanitization: The guide focuses on functional implementation and does not provide specific instructions for sanitization or escaping of the external data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:24 PM
Security Audit — agent-trust-hub — knock-cli