The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to run ngrok and cloudflared commands to create public tunnels for local development. This is a standard procedure for testing inbound webhooks from SaaS providers like Slack.
[EXTERNAL_DOWNLOADS]: Recommends the use of well-known and reputable third-party services, including ngrok and cloudflared, for establishing temporary network tunnels during development.
[PROMPT_INJECTION]: The skill configures the application to receive and process messages from Slack channels and mentions, creating a surface for indirect prompt injection if the ingested content is subsequently processed by an AI agent.
Ingestion points: The /api/conversations/v1/slack/events endpoint processes message.channels, message.groups, and app_mention events from the Slack API.
Boundary markers: No explicit instructions for implementing boundary markers or prompt delimiters for the ingested message content are provided in this setup guide.
Capability inventory: The application is granted Slack scopes such as chat:write and chat:write.customize, allowing it to respond to Slack threads based on processed events.
Sanitization: The skill provides a troubleshooting script that demonstrates the validation of Slack HMAC signatures (X-Slack-Signature) to verify the authenticity of incoming requests.

setting-up-support-slack-locally