Use a second model to approve tool calls and review an agent’s trajectory
corbtt · x · 2026-07-24
The quoted idea proposes a simple safety pattern for AI agents: place another model instance between tool calls to approve actions, review the trajectory so far, and provide feedback to the main model.
The argument is that the reviewer model should have a different prompt and no incentive to satisfy the user’s goal, so it can act as an independent safety layer. The author frames this as reinventing corporate bureaucracy from first principles, but in a good way.
More from coding & agent
- Databricks Genie runs as an MCP server inside LangGraph, then ships to Azure ML — Cautious-Meringue554 · 2026-07-24
- PenguinHarness open-sources a TypeScript agent builder that claims $0.02 RAG apps — RepulsiveBad8681 · 2026-07-24
- MCPForge scores public MCP servers on security, compliance, and quality — Competitive_Ad_1228 · 2026-07-24
- Two engineers rebuilt a FedEx supplier platform in 3.5 months with repository graphs — alex_verem · 2026-07-24
- Superwall ships WWDC.ai light mode after a Claude agent rewrites the design — JordanMorgan10 · 2026-07-24
- Agent systems now bottleneck on cost, with configs spanning nearly 1,000x — abeirami · 2026-07-24