# Slack (/context-providers/providers/slack) Two tools: `query_slack` for searching and reading, `update_slack` for posting messages. Complete [Setup](#setup) before running this example. Search needs a user token in addition to the bot token. Save as `slack_context.py` and run `python slack_context.py`: ```python title="slack_context.py" import asyncio from agno.agent import Agent from agno.context.slack import SlackContextProvider from agno.models.openai import OpenAIResponses async def main(): slack = SlackContextProvider(write=False, model=OpenAIResponses(id="gpt-5.4-mini")) agent = Agent( model=OpenAIResponses(id="gpt-5.4"), tools=slack.get_tools(), instructions=slack.instructions(), ) await agent.aprint_response("What did the team discuss about the auth migration?") if __name__ == "__main__": asyncio.run(main()) ``` The agent can call `query_slack`; its sub-agent searches and reads relevant threads. This example exposes no posting tool. ## When to use this vs SlackTools [#when-to-use-this-vs-slacktools] | Use SlackContextProvider when... | Use SlackTools directly when... | | --------------------------------------------------------- | ----------------------------------------------- | | Slack is one of several context sources | Slack is the primary task surface | | You want a small source-level tool surface | You need fine control over individual API calls | | The agent should ask questions, not orchestrate API calls | You're building a Slack-specific agent | ## Setup [#setup] In a [virtual environment](/sdk/setup), install the SDKs and configure your Slack app tokens: ```shell uv pip install -U agno openai slack-sdk aiohttp export OPENAI_API_KEY="your-openai-api-key" export SLACK_USER_TOKEN=xoxp-... export SLACK_BOT_TOKEN=xoxb-... ``` Or pass directly: `SlackContextProvider(token="xoxb-...")`. Install the Slack app in your workspace with the scopes needed for search and channel/thread reads. Invite it to private channels it must read. With only a bot token, ask about a specific accessible channel or thread; full-text search is unavailable. See [SlackTools](/tools/toolkits/social/slack#toolkit-params) for OAuth scope requirements. ## Example queries [#example-queries] Queries that work well give the provider a topic to search and context to narrow results: | Query | Why it works | | -------------------------------------------------------------- | ---------------------------------- | | "What did the team decide about billing migration?" | Topic-driven search with synthesis | | "Summarize the thread where Alex discussed OAuth scopes" | Combines search + thread reading | | "What open questions came up in #launch this week?" | Channel + time constraints | | "Post a short update to #support saying the issue is resolved" | Clear channel + action | Queries that are too vague: | Less effective | Better | | ---------------- | ---------------------------------------------------- | | "Search Slack" | "Search for recent discussion about webhook retries" | | "What happened?" | "What happened in #incidents about the API outage?" | | "Post it" | "Post this summary to #engineering" | ## Configuration [#configuration] | Parameter | Type | Default | Description | | -------------------- | ------------- | --------- | ------------------------------------------------------------------------- | | `id` | `str` | `"slack"` | Changes tool names to `query_` and `update_`. | | `model` | `Model` | `None` | Model for the sub-agents. Defaults to Agno's default model. | | `read` | `bool` | `True` | Expose `query_slack`. | | `write` | `bool` | `True` | Expose `update_slack`. | | `enable_media_tools` | `bool` | `False` | Enables `download_file` for read tools and `upload_file` for write tools. | | `mode` | `ContextMode` | `default` | `default` exposes both tools. `tools` exposes read-only SlackTools. | ## Read/write modes [#readwrite-modes] Control what the agent can do: ```python # Research agent: can search but not post slack = SlackContextProvider(write=False) # Notification agent: can post but not search slack = SlackContextProvider(read=False) ``` Use `write=False` for research, triage, audit, and eval agents. The read sub-agent physically cannot post because `send_message` isn't in its toolkit. ## Multi-provider example [#multi-provider-example] After completing [Google Drive setup](/context-providers/providers/drive) too, replace the body of `main()` with this composition fragment. Add the imports at module scope: ```python from agno.context.slack import SlackContextProvider from agno.context.gdrive import GoogleDriveContextProvider slack = SlackContextProvider(write=False) drive = GoogleDriveContextProvider() agent = Agent( model=OpenAIResponses(id="gpt-5.4"), tools=[*slack.get_tools(), *drive.get_tools()], instructions="Use Slack for team discussion, Drive for docs. Note when they disagree.", ) await agent.aprint_response("What's the current auth spec and did engineering raise concerns?") ``` The agent sees 2 tools (`query_slack`, `query_gdrive`) on the calling agent. ## Tips [#tips] * **Token naming**: `SLACK_BOT_TOKEN` is preferred; `SLACK_TOKEN` is a fallback. * **Search**: `search_messages` needs a user token. Set `SLACK_USER_TOKEN` or pass `user_token=`. With only a bot token, the read sub-agent uses channel history and threads. * **Private channels**: The bot must be invited to access private channel history. * **Thread context**: Ask for "the thread" or "the decision" to read replies rather than only the parent. The current `get_thread` call defaults to at most 20 messages and does not paginate. * **Channel names**: Natural names like `#general` work. The sub-agent resolves them to IDs. * **Write clarity**: "Post this to #team" is safer than vague "send it." ## Resources [#resources] Toolkit methods and OAuth scope requirements Working example with search and posting