Indent

What is Indent?

Indent is a shared AI coworker your whole company uses from Slack, GitHub, or the web. Hand it a bug, a pull request, or a data question, and it works on a real computer, running commands, editing files, and querying databases. What one person teaches it, everyone benefits from.

Top Features:

• Slack tasks: mention it in any channel and assign work in plain language.
• Code review: checks pull requests against your own plain-text rules and suggests autofixes.
• Data answers: queries PostgreSQL, BigQuery, or ClickHouse and returns answers and reports.

Use Cases:

• Incident triage: trace a production alert through Datadog or Sentry and fix it.
• Self-serve analytics: let product and finance staff get data answers without engineer help.
• Ticket handoff: assign a Linear issue and get back a reviewable pull request.

Who Can Use Indent?

• Engineering teams: offload bug fixes, reviews, and on-call investigations to one shared helper.
• Product managers: get quick answers from company data without filing analyst requests.
• Startup CTOs: replace separate code review and coding tools with one learning assistant.

Pricing

• Pay as you go (from $5): buy credits in any amount and pay only when work runs.
• Team (usage-based): no seats, so the whole company shares one coworker and balance.
• Enterprise (contact sales): invoiced monthly credit grant with SSO, security reviews, and a DPA.

Pros and Cons

Pros:

• No seat fees: adding people costs nothing since you only pay for work done.
• Shared memory: rules one teammate teaches it apply across the whole company.
• Shows its work: each task lists the commands run and how results were verified.

Cons:

• Mac only locally: the desktop app needs macOS on Apple Silicon for local sessions.
• No free tier: work stops once the trial ends until someone adds paid credits.
• Setup effort: best results need GitHub, a dev environment, and data tools connected.

FAQs:

1) Does it work inside Slack?

Yes, invite it to a channel and mention it with a task.

2) How is it billed?

Usage-based credits with no plans or seats, plus admin spending limits.

3) Can it open pull requests?

Yes, it edits code, runs tests, and opens a tested pull request.

4) Which data sources can it query?

PostgreSQL, BigQuery, ClickHouse, plus Datadog, Sentry, and Notion through MCP connections.

5) Does it run locally?

Yes, a macOS desktop app runs local sessions, alongside cloud sessions.