Skip to content
Skip to contents
Skip to reading controls

Search skills⌘K

Setting the file. One moment.

This file

Type

Size

Lines

Position

prisma
/
prisma/skills

Agent Safety · Prisma CLI · prisma/skills · Skills Docs

Repo No. 95 · Prisma CLI

↖Back to the cover

Skills · Prisma…

Prisma CLI
Prisma Client API
Prisma Compute
Prisma Database Setup
Prisma Driver Adapter Implementation
Prisma MongoDB Upgrade
Prisma Postgres Setup
Prisma Postgres
Prisma Upgrade V7

9 skills · 29 min

Skill 01 · Prisma CLI

Subchapter 1.1

Agent Safety

references/agent-safety.mdMarkdown1 KBView on GitHub

AI safety checkpoint for destructive commands

Prisma detects common AI-agent environments and blocks these commands until the user gives explicit consent:

prisma migrate reset
prisma db push --force-reset
prisma db push --accept-data-loss

1Up to skill 1Prisma CLI

On this page

AI safety checkpoint for destructive commands
Required workflow
Reference

This file

Number: 1.1
Position: 1 of 20
Type: Markdown
Size: 1 KB

Nearby

Next Complete

Source

references/agent-safety.md

Required workflow

Inspect the target database/config and explain exactly what can be deleted or reset.
Ask the user for explicit consent immediately before the action.
Run the command only after that consent.

For an agent-run subprocess, Prisma accepts the exact consent text through:

bash

PRISMA_USER_CONSENT_FOR_DANGEROUS_AI_ACTION='<exact user consent message>' prisma migrate reset --force

The value must match the user’s message exactly and must not contain added quotes or newlines. Never fabricate consent, reuse an old unrelated approval, or bypass the checkpoint by hiding agent-detection environment variables.

The MCP server has no migrate-reset tool. Use the shell command only after consent.

Reference

Prisma ORM 7.9.0 release (opens in a new tab)

First file of the skillBack to Prisma CLIReferenceComplete

(opens in a new tab)

Raw file(opens in a new tab)

prisma/skills(opens in a new tab)