The Agent Skills Directory

[COMMAND_EXECUTION]: The script executes the slack-post CLI tool and a sibling secrets helper script located at ../../secrets/scripts/secrets to manage its environment and post messages.\n- [CREDENTIALS_UNSAFE]: The skill manages sensitive credentials including SLACK_TOKEN and SLACK_BOT_TOKEN stored in the local file ~/.secrets/.env.slack. It includes explicit safety instructions to prevent the agent from printing or echoing these secret values in the chat.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided text to send as Slack notifications, which creates a surface for indirect prompt injection. A malicious input could attempt to bypass constraints or misrepresent the source of a message.\n
Ingestion points: The slack-notify script accepts message content via command-line arguments and standard input in scripts/slack-notify.\n
Boundary markers: The instructions specify confirming user intent but do not define specific delimiters or 'ignore embedded instructions' markers for the message payload.\n
Capability inventory: The skill can execute local scripts, invoke the slack-post tool, and read environment variables via the secrets helper.\n
Sanitization: User arguments are handled using standard shell quoting to prevent local command injection, but the message content is passed directly to the notification tool without sanitization for downstream platforms.

slack-notify