The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bash script (scripts/sw-editor.sh) to execute curl for network requests and jq for JSON parsing and construction.\n- [EXTERNAL_DOWNLOADS]: The CLI wrapper communicates with Superwall's official relay API at https://superwall-mcp.superwall.com and https://superwall.com to manage browser editor sessions.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes tool definitions and paywall data received from the browser editor. This creates a surface where content within a paywall could influence agent behavior. \n
Ingestion points: Tool definitions and results fetched via scripts/sw-editor.sh tools and call.\n
Boundary markers: Absent; the skill relies on the browser session as the source of truth for tool schemas.\n
Capability inventory: The CLI performs network requests and shell operations; the agent can modify paywall HTML, styles, and logic.\n
Sanitization: Arguments are safely embedded into JSON payloads using jq --argjson.\n- [SAFE]: The skill implements security best practices by restricting state file permissions to the current user (chmod 600) and providing a release command to securely clear session tokens and local state.

superwall-editor