Parseable

Slackbot

Ask Parseable questions from Slack using Slackbot and MCP.


The Parseable Slackbot brings Parseable investigation into the Slack conversation where an incident, alert, or debugging thread is already happening. Instead of switching context to copy a query into a browser, a user can connect their Parseable account from Slack and ask questions with @parseable.

The bot is available for Parseable OSS, Cloud, and Enterprise users. There is no separate setup flow for each edition. Slackbot works through the Parseable MCP path and is accessed from Slack chat. Slack is the place where the user asks the question, MCP provides the tool interface, and Parseable returns data from logs, metrics, traces, alerts, and other supported resources.

How It Works

Slack user
  |
  | /parseable connect
  | API key
  v
Parseable Slackbot
  |
  | MCP request
  v
Parseable MCP Server
  |
  | Parseable API request
  v
Parseable Cloud, Enterprise, or OSS

The bot does not ingest telemetry or create a new data store. It calls Parseable through MCP and returns the answer in Slack. Parseable still authenticates the API key and enforces the permissions attached to that key.

This works across Parseable Cloud, Parseable Enterprise, and self-hosted OSS deployments. The exact actions available in Slack depend on the Parseable API key used during connection.

What You Can Ask

Once connected, mention the bot in a Slack channel or thread and ask a question:

@parseable show me errors from api-gateway in the last 10 minutes

You can use the bot for the same kind of investigation you would normally do in Parseable or through the MCP Server:

  • Find datasets and inspect schema.
  • Query logs with a bounded time range.
  • Run PromQL for metrics.
  • Look up traces and service behavior.
  • Inspect alerts and alert targets.
  • Review users, roles, and access information when the API key allows it.
  • Create or evaluate alerts when the connected key has those permissions and the action is confirmed.

For read-only workflows, connect with a read-only Parseable API key.

Prerequisites

Before you start, make sure you have:

  • Access to the Slack workspace where the Parseable bot is installed.
  • A Parseable Cloud, Enterprise, or OSS deployment.
  • A Parseable API key with the permissions you want the bot to use.

Connect From Slack

Run the connect command in Slack:

/parseable connect

The bot asks for the details it needs to reach Parseable. For Parseable Cloud, provide the API key for the workspace you want to query. For Enterprise or OSS, provide the Parseable URL and API key if the bot asks for both.

After the connection is saved, ask a small test question:

@parseable list datasets

If the bot returns the dataset list, the Slack workspace is connected to Parseable and ready for investigation.

Ask Questions in Context

The bot is most useful when the question already has context. You can ask inside an alert thread, an incident channel, or a service-specific channel.

Examples:

@parseable what changed in checkout errors during the last 30 minutes?
@parseable query payment-api logs for 5xx errors in the last hour
@parseable show request latency for frontend over the last 15 minutes
@parseable summarize recent alerts for production

Keep the first question narrow. Start with a dataset, service name, error type, or time window. Once the bot finds the right direction, ask follow-up questions in the same thread.

Permissions and Safety

Slackbot follows the same security model as the MCP Server. It does not get special access to Parseable. If the connected API key cannot access a dataset, the bot cannot access it either.

Use these rules for production work:

  • Use least-privilege API keys.
  • Prefer read-only keys for investigation channels.
  • Use a separate key for workflows that need alert creation or alert evaluation.
  • Rotate the key if it is shared in the wrong channel or exposed outside Slack.
  • Avoid pasting API keys in public channels.

Slack permissions are separate from Parseable permissions. Slack scopes control what Slackbot can do inside Slack. The connected Parseable API key controls what it can access inside Parseable.

Slackbot uses the following bot token scopes:

OAuth scopeDescription
app_mentions:readView messages that directly mention @Parseable in conversations that the app is in.
channels:historyView messages and other content in public channels that Parseable has been added to.
channels:readView basic information about public channels in a workspace.
chat:writeSend messages as @Parseable.
commandsAdd slash commands, including /parseable connect.
groups:historyView messages and other content in private channels that Parseable has been added to.
im:historyView messages and other content in direct messages that Parseable has been added to.
mpim:historyView messages and other content in group direct messages that Parseable has been added to.
team:readView the name, email domain, and icon for workspaces that Parseable is connected to.
users:readView people in a workspace.
users:read.emailView email addresses of people in a workspace.

If the app is submitted to the Slack Marketplace, Slack reviews the requested scopes before approval. After approval, Slackbot can only use the scopes Slack has approved for that app.

How This Differs From Slack Alert Targets

Slackbot is for asking questions and receiving answers in Slack.

Slack alert targets are different. They send alert notifications from Parseable to a Slack channel when an alert fires. If you only want alert messages in Slack, use the Slack alerting integration. If you want to ask questions from Slack, use Slackbot.

Both can be useful together. An alert can arrive in Slack, and the same thread can continue with an @parseable question to inspect the logs, metrics, traces, or alerts behind the incident.

Was this page helpful?

On this page