service-digital-engagement-channel-configure

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill workflow involves executing shell commands using the sf (Salesforce CLI) and local bash scripts to query metadata and deploy configurations to a Salesforce organization. These operations are within the scope of the skill's intended purpose as a developer tool.
  • [PROMPT_INJECTION]: The skill ingests user-supplied data such as channel labels and routing targets which are interpolated into shell commands (sf data query, sf data update) and local scripts. This represents a potential surface for indirect injection if inputs are not properly escaped by the agent.
  • Ingestion points: Channel label, routing target, and fallback queue names collected from user prompts in SKILL.md.
  • Boundary markers: Absent; the instructions do not explicitly mandate delimiters or escaping for user-supplied strings when interpolated into commands.
  • Capability inventory: The skill possesses the capability to query the org, deploy metadata files, and update records via the Salesforce CLI.
  • Sanitization: The skill includes a dedicated script scripts/normalize-channel-name.sh that uses sed to sanitize the channel label, ensuring it conforms to API naming conventions (alphanumeric and underscores only).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 09:29 PM
Security Audit — agent-trust-hub — service-digital-engagement-channel-configure