The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents an extensive set of commands for the suprsend CLI, allowing the agent to manage categories, workflows, events, and schemas. This includes the ability to start an MCP server and generate type definitions.
[CREDENTIALS_UNSAFE]: The tool manages authentication through service tokens. It provides the suprsend profile add command to store tokens in a local configuration file ($HOME/.suprsend.yaml) and supports the SUPRSEND_SERVICE_TOKEN environment variable for authentication. These are standard CLI patterns for credential management.
[EXTERNAL_DOWNLOADS]: The CLI tool performs network operations to SuprSend's official API and management endpoints (e.g., management-api.suprsend.com, hub.suprsend.com) to synchronize notification assets, schemas, and translations.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it retrieves data from an external SuprSend workspace that the agent is then expected to process or act upon.
Ingestion points: Data enters the context via suprsend workflow pull, suprsend category pull, suprsend event pull, and suprsend schema pull commands.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions potentially embedded in the descriptions or metadata of the retrieved assets.
Capability inventory: The agent can modify workspace states, commit changes, and write to the local filesystem using commands like push, commit, pull, and generate-types across all asset categories.
Sanitization: The documentation does not specify any sanitization or validation of the content (such as node descriptions or schema properties) before it is presented to the agent.

suprsend-cli