The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes Ruby scripts via the shell to interact with the PostHog API.- [DATA_EXFILTRATION]: The skill transmits the PostHog Personal API Key to the official PostHog API (app.posthog.com or a self-hosted instance specified by the user). This is the intended behavior for the management of project data.- [CREDENTIALS_UNSAFE]: During setup, the skill accepts the API token as a command-line argument in save_token.rb. The token is then persisted in ~/.config/posthog/token with 0600 permissions, which is a standard practice for local CLI tools to protect secrets from other users on the system.- [PROMPT_INJECTION]: The skill displays data fetched from the PostHog API, such as event names and person identifiers, which presents a surface for indirect prompt injection.
Ingestion points: Data is ingested from PostHog API responses in events.rb, persons.rb, insights.rb, projects.rb, and feature_flags.rb.
Boundary markers: No delimiters are used to wrap the external data in the output.
Capability inventory: The skill can modify feature flag statuses via toggle_flag.rb and the auth.rb module provides a helper for POST, PUT, and DELETE requests.
Sanitization: Content retrieved from the API is displayed without filtering or sanitization.

posthog