INTELLIGENCE
Experts Inference Models Pricing Skills Playbooks Articles Resources
Sign in Meet your experts
  1. Home
  2. /Articles
  3. /AI agents
  4. /Agent safety and permissions
  5. /Dry runs and idempotency

Dry runs, confirmation, and idempotency

Tool scope and permissionsPrompt injection defenseDry runs and idempotencyReversibility and undoKill switches and denials
AI agents

Dry runs, confirmation, and idempotency

  • A Dry-Run Endpoint for Reversals: Testing the Undo First
  • Idempotency Keys for Agent Writes Behind a Human Gate
  • Two-Phase Writes From the Agent's Side: Dry Run, Then Commit
  • A Commit Token That Carries the Plan: Fields and a TTL
  • A Confirmation Token the Model Cannot Forge
  • A Read-Only Rehearsal of a Write Tool Before It Fires
  • Choosing Between JSON Schema and Pydantic for Agent Contracts
  • Contract Tests for a Dry-Run Endpoint
  • Designing Idempotency Keys for Agent Retries
  • Idempotency Keys in Practice: A Design Checklist for Write Tools
  • Idempotency Keys on the Undo: Proving It Ran Exactly Once
  • Idempotent Writes Behind an Approval Gate: No Double Runs
  • Write Tools: Idempotency Keys and the Double Refusal
Agent safety and permissions

Experts · Inference · Models · Docs · Skills · Pricing · Articles · Skills API · About

© 2026 Quanta Intelligence · 8 The Green STE B, Dover, DE 19901