The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The CLI manages authentication via service tokens, which are stored in a local configuration file at ~/.suprsend.yaml. The suprsend profile subcommands are specifically designed to store and modify these sensitive credentials.
[DATA_EXFILTRATION]: Multiple commands, such as suprsend workflow push and suprsend category push, transmit local configuration and asset data to SuprSend's external APIs (hub.suprsend.com and management-api.suprsend.com).
[COMMAND_EXECUTION]: The skill documents the generation and execution of shell completion scripts and type definition generators for various programming languages, which are used to integrate the tool into local development environments.
[PROMPT_INJECTION]: The suprsend start-mcp-server command facilitates AI agent interaction with SuprSend resources, creating a potential surface for indirect prompt injection.
Ingestion points: Data retrieved from the SuprSend workspace (e.g., notification templates, event schemas, documentation) via MCP tools like documentation.fetch or workflow.get.
Boundary markers: The documentation does not indicate the use of specific delimiters or instructions to prevent the agent from obeying instructions embedded in the retrieved data.
Capability inventory: The CLI has the ability to pull and push workspace assets, manage user profiles, and generate executable code snippets.
Sanitization: There is no documented sanitization process for data fetched from the remote service before it is provided to the agent.

suprsend-cli