The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from various external ad platforms and analytics tools which may contain malicious instructions designed to influence the agent.
Ingestion points: Reporting data from Google Ads, Meta Ads, TikTok, GA4, and PostHog accessed via soku ads, soku ga4, and soku posthog namespaces.
Boundary markers: No specific delimiters or safety warnings are provided to prevent the agent from obeying instructions embedded in external analytics data.
Capability inventory: Access to network APIs, advertising campaign management, file creation/publishing via SEO hosting, and subprocess execution via exec().
Sanitization: The skill does not describe any sanitization, validation, or filtering of the external data before it is presented to the agent context.
[COMMAND_EXECUTION]: The provided integration examples in TypeScript and Shell demonstrate unsafe patterns for executing system commands.
Evidence: The TypeScript code snippet uses template literals for command assembly: `soku ads list-ad-accounts --platform ${platform}`. This pattern allows for command injection if the platform variable is sourced from untrusted user input.
[EXTERNAL_DOWNLOADS]: The skill requires the installation and use of an external CLI package from the npm registry.
Evidence: The installation section explicitly calls for npm install -g @soku-ai/cli and npx @soku-ai/cli --help to fetch and run remote code.

soku-cli-integration