Slack Summary Reporter

A simple Python Actor that fetches recent Slack messages, summarizes them via OpenAI, and posts the summary back to Slack—configured entirely via per-run JSON input instead of environment variables.

🛠️ Prerequisites

• Python 3.12+
• Docker (optional, for containerized runs)
• An Apify account and the Apify CLI installed (for local development)
• A Slack app with OAuth scopes:
  • conversations.history (to read messages)
  • chat:write (to post messages)
• An OpenAI API key

⚙️ Actor Input

All configuration now comes via the Actor’s INPUT schema (.actor/actor.json). Users supply a JSON object at run-time—either in the Apify Console form, as a Task, or via apify run --input input.json.

{
  "summaryDays": 1,                      // How many days back to fetch
  "slackTokenFetch": "xoxb-…",           // [secret] token with history scope
  "slackChannelFetch": "C0123456789",    // Channel ID to read from
  "slackTokenPost": "xoxb-…",            // [secret] token with chat:write scope
  "slackChannelPost": "C9876543210",     // Channel ID to post summary
  "openaiApiKey": "sk-…",                // [secret] your OpenAI API key
  "openaiModel": "gpt-4o-mini"           // Which OpenAI model to use
}

Note: Fields marked secret (slackTokenFetch, slackTokenPost, openaiApiKey) are defined with isSecret: true in the Actor schema, so they’re encrypted in the Console and API.

🚀 Running Locally

1. Push your Actor (includes .actor/actor.json) to Apify:

   $apify push
   

2. Create a local input.json (see example above).
3. Run with the Apify CLI, pointing to your input file:

   $apify run --input input.json
   

This will execute main.py under the Apify runtime, loading your JSON as await Actor.get_input().

🐳 Docker Usage

To build and run the container locally, mount your input.json into the Apify default key–value store:

1. Build the image:

   $docker build -t slack-summary-reporter .
   

2. Run with volume mounts for input storage:

   # Create a folder for the default store
   mkdir -p apify_storage/key_value_stores/default
   # Copy your input.json there
   cp input.json apify_storage/key_value_stores/default/INPUT.json
   # Run the container
   docker run --rm      -v "$(pwd)/apify_storage":/home/apify_user/apify_storage      slack-summary-reporter
   

The Actor will read /home/apify_user/apify_storage/key_value_stores/default/INPUT.json automatically.

🔧 Customization

• summaryDays: Adjust how many days of history to include.
• summaryPromptTemplate (optional): Override the default summarization prompt.
• maxChunkTokens (optional): Change the OpenAI-chunk-size limit.

All of these are additional properties you can add to your .actor/actor.json schema and supply in the same input JSON.

📖 License

MIT © Marko Kubrachenko