accidental-data-loss-prevention

Requires explicit user consent before executing irreversible data deletion commands.

1|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/tottenjordan/me-skittles --skill accidental-data-loss-prevention-tottenjordan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: accidental-data-loss-prevention
Source: https://github.com/tottenjordan/me-skittles/tree/main/gemini/accidental-data-loss-prevention
Command: npx skills add https://github.com/tottenjordan/me-skittles --skill accidental-data-loss-prevention-tottenjordan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI agents can accidentally run destructive commands like DROP TABLE, TRUNCATE, or gsutil rm that permanently delete production data. This Skill enforces a mandatory stop-and-verify checkpoint so no irreversible deletion happens without explicit user approval. ## Core Features & Use Cases - Destructive SQL Guard: Intercepts DROP TABLE/VIEW/SCHEMA/DATABASE, TRUNCATE, and broad DELETE statements missing a WHERE clause or using 1=1. - Cloud Resource Protection: Blocks unconsented gsutil rm or gcloud storage rm on production buckets, plus deletion of Spanner, BigQuery, Dataproc resources, secrets, and KMS keys. - Mandatory Consent Workflow: Halts execution, explains the impact and rationale, and waits for explicit affirmative approval before proceeding. - Use Case: While cleaning up a database, the agent proposes running DROP TABLE on a production schema; the Skill forces it to pause, explain the impact, and wait for your confirmation first. ## Quick Start Ask the agent to clean up old database tables and it will pause to request your explicit approval before running any destructive deletion command.

Frequently Asked Questions about accidental-data-loss-prevention

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent an AI agent from deleting production data?▼

Use a consent-gating skill that halts any irreversible deletion command before execution. The agent must explain the impact, justify why deletion is needed, and wait for your explicit affirmative approval before proceeding.

What SQL commands are blocked by accidental data loss prevention?▼

The Skill intercepts DROP TABLE, DROP VIEW, DROP SCHEMA, DROP DATABASE, TRUNCATE, and broad DELETE statements that lack a WHERE clause or use 1=1. These commands require explicit user consent before execution.

Does this protect Google Cloud Storage buckets from deletion?▼

Yes, gsutil rm and gcloud storage rm commands targeting production data or critical buckets require explicit consent. It also covers deletion of Spanner, BigQuery, Dataproc resources, secrets, and KMS key destruction.

When should I use a data loss prevention guardrail?▼

Use it whenever an AI agent has access to production databases, cloud storage, or infrastructure resources. It is most valuable during cleanup, migration, or maintenance tasks where destructive commands are plausible.

What happens when the agent proposes a destructive command?▼

The agent halts execution, explains the impact of the deletion and why it believes the action is necessary, then requests explicit approval. It only proceeds after receiving clear affirmative consent in the conversation.