Escanor

Use case

Automate operations with Autopilot

Autopilot works toward a goal you write, across the tools you connected, and records every step. Risky actions wait for a person. You set the autonomy level and the budgets, and one switch pauses everything.

Sign up freeRead the automations guide

How it works

  1. 01

    Goal

    What to do, and what “done” looks like.

  2. 02

    Run

    Steps across your connected tools, each one recorded.

  3. 03

    Needs you

    An owner or manager approves, denies or stops it.

  4. 04

    Verify

    Check the result against your criteria.

  1. Step 1

    Write a goal

    Open Automations, enter a goal and optional “Done when” criteria, and start it. A good first goal is read-only, on staging.

    Start and inspect a run

  2. Step 2

    Follow the run

    Each run records its steps and decisions and keeps going after you leave the page. Its status reads Running, Needs you, Succeeded, Failed or Stopped.

    Automations and Autopilot

  3. Step 3

    Decide on what matters

    Anything that needs a decision waits under Needs you. Approve the action you understand, or deny or stop it.

    Approvals for automations

  4. Step 4

    Add triggers

    Set up triggers and multi-step missions with a goal, criteria and a cooldown. Start a trigger disabled, test it, then turn it on.

    Triggers and rules

A first goal to try

Goal: Investigate the most recent failed staging deployment.
Done when: The report identifies the first build error, links the evidence,
and proposes one fix. Do not deploy or change configuration.

You set the limits

Autonomy level
Decide how much Autopilot may do on its own, with per-action overrides.
Budgets
Cap runs per day and the number of automatic approvals.
Verification and notifications
Choose how results are checked and who hears about them.
Stop everything now
Pause all automation in the workspace with one switch, then resume when you are ready.

Auto-fix for incidents

Auto-fix lists errors that can be repaired, alongside repair missions and playbooks. Each proposed fix shows its target, branch, commands and tests, and waits for approval before it changes anything. See AI incident response.

Start runs from your own tools

A signed-in session can start a run with POST /api/v1/autopilot/runs, then approve, deny or stop it through the same API.

{"goal":"Investigate the staging build failure without changes","criteria":"Report the first error and evidence"}

AI runs on every plan

Free includes 50 AI runs a month, Pro 1,000 and Team 10,000. Enterprise limits are set to fit your team. See pricing.

Sign up freeWatch the demoBook a call