Lifecycle and Support

Recovery & Offboarding

What It Means

Sometimes you might need to remove an AI Employee from your workspace or migrate to a different one.

When This Happens

  • You decide you no longer need an AI Employee

  • Motion retires an AI Employee (stops supporting it)

  • You want to switch to a different AI Employee

  • You're leaving your workspace and need to hand off AI Employees to someone else

What Happens When You Remove an AI Employee

  • The AI Employee is removed from your workspace

  • All skills associated with that AI Employee stop working

  • Any autonomous skills that were running will no longer run

  • The AI Employee no longer appears in your workspace directory

How to Migrate to a Different AI Employee

If you want to switch to a different AI Employee:

  1. Identify which skills from the old AI Employee you still need

  2. Create similar skills with the new AI Employee

  3. Update any triggers or workflows that used the old AI Employee

  4. Test the new skills to make sure they work

  5. Remove the old AI Employee once you're confident the new one is working

What You Should Do

  • If an AI Employee is being retired, start planning your migration early

  • Document your current skills before removing an AI Employee

  • Test new skills thoroughly before removing the old ones

  • Keep notes on how to recreate important skills in case you need to migrate


Troubleshooting

AI Employees may encounter problems when running skills or using integrations. This section lists common issues and how to fix them.

General Troubleshooting Approach

When something goes wrong:

  1. Check the skill's run history to see what error occurred

  2. Read the error message carefully — it usually tells you what went wrong

  3. Use the "Common Issues and Fixes" list below to find the solution

  4. If the fix doesn't work, try the prevention tips to avoid the problem in the future

Common Issues and Fixes

  1. Issue: "Permission Denied" Error

What This Means

The AI Employee tried to do something that the context user doesn't have permission to do.

Why It Happens

  • The context user is a Member (not an Admin) and tried to perform an Admin-only action

  • The context user doesn't have access to the workspace or project

  • The context user's role was recently downgraded

How to Fix It

  • Check what role the context user has (Admin or Member)

  • If the action requires Admin permissions, upgrade the user to Admin

  • If the context user needs access to a workspace or project, grant them access

  • If the action is Admin-only and you don't want to upgrade the user, change the skill to perform a different action that Members can do

Prevention

  • Review trigger rules regularly to make sure they still make sense

  • Keep notes on which actions require which permissions

  • Test skills manually before setting them to run automatically

  1. Issue: Skill Fails With External Service

What This Means

The skill tried to use an external service (like Gmail, Slack, or Salesforce) but couldn't connect.

Why It Happens

  • The connection hasn't been set up yet

  • The connection has expired (you changed your password, for example)

  • The external service is temporarily unavailable

  • The connection was revoked

How to Fix It

  • Check if the connection is set up in your workspace settings

  • If the connection exists but is expired, reconnect it by authenticating again

  • If the external service is down, wait a few minutes and try again

  • If the connection was revoked, set it up again

Prevention

  • Monitor audit logs to catch connection failures early

  • Revoke unused connections so you don't have old expired connections lying around

  • Test skills manually before setting them to run automatically

  • Keep notes on which connections each skill uses

  1. Issue: Action Blocked for Approval (Manual Mode Only)

What This Means

The AI Employee tried to perform a write action (create, edit, or delete) in manual mode, and you need to approve it before it can proceed.

Why It Happens

  • This is normal behavior in manual mode — the AI Employee always asks for approval before write actions

  • You haven't clicked "Approve" yet

How to Fix It

  • Review what the AI Employee is about to do

  • If it looks correct, click "Approve"

  • If it doesn't look right, click "Decline" and ask the AI Employee to try again

Prevention

  • This isn't really a problem — it's a safety feature

  • Just review the action and approve it if it looks correct

  1. Issue: Autonomous Skill Not Running

What This Means

A skill that's supposed to run automatically (on a schedule or when an event happens) isn't running.

Why It Happens

  • The trigger is misconfigured

  • The trigger event isn't happening (for example, you set it to run when a new task is created in Project A, but no one is creating tasks in Project A)

  • The skill is outside its defined scope

  • The connection the skill needs is missing or revoked

How to Fix It

  • Review the skill configuration to make sure the trigger is set up correctly

  • Check if the trigger event is actually happening (for example, are tasks actually being created?)

  • Make sure the skill has access to the connections it needs

  • Reconnect any missing connections

Prevention

  • Test skills manually before setting them to run automatically

  • Review trigger rules regularly to make sure they still make sense

  • Monitor audit logs to catch skills that stop running

  1. Issue: Output Missing From Project or Doc

What This Means

You expected the AI Employee to create something (like a task or a doc), but it's not there.

Why It Happens

  • The action was declined (you didn't approve it in manual mode)

  • The action was deleted after the AI Employee created it

  • The connection failed mid-run

  • The skill ran but encountered an error

How to Fix It

  • Check the skill's run history to see if the action was completed

  • If it shows as completed, check if it was accidentally deleted

  • If it shows as failed, read the error message to understand what went wrong

  • Re-run the skill if needed

Prevention

  • Check the skill's run history regularly to make sure it's working

  • Don't accidentally delete outputs from AI Employees

  • Monitor audit logs to catch failures early

  1. Issue: Failed Skill Run Always Shows Why It Failed

What This Means

When a skill fails, Motion always logs the reason why. You can see exactly what went wrong.

Where to Find the Error

  • Check the skill's run history

  • Look for the failed run and click on it

  • Read the error message — it will tell you what went wrong

  • The error message might also tell you which step failed

Common Error Codes

  • 400 — Bad request (usually means the data format is wrong)

  • 401 — Authentication failed (usually means the connection expired)

  • 403 — Permission denied (usually means you don't have access)

  • 404 — Not found (usually means the thing you're trying to access doesn't exist)

  • 500 — Server error (usually means the external service is having issues)

What You Should Do

  • Read the error message carefully

  • Use the error code and message to figure out what went wrong

  • Use the troubleshooting table above to find the fix

  • If you're still stuck, check Motion's help documentation or contact support

Last updated

Was this helpful?